Ian --

I believe the theory of heating at the top is to achieve an uninterrupted 
gradient from the top to the bottom of the enclosure.  (Naturally, this 
cannot ever totally be achieved but one can come close.  There will 
necessarily be convection currents in the vicinity of the heater but 
hopefully confined to that area.)  If the gradient can be maintained at all 
times, there should be minimal convection currents since there is never air 
of lower density below air of higher density (which would want to make the 
air change places via convection).  A requirement is that any outside heat 
flowing into the box below the heater be at a low enough rate that the 
gradient is never reversed (never gets to a point of having cooler air 
above warmer).  Similarly, heat flowing out of the box near the top must be 
slow enough to prevent reversing the gradient.  This requires insulation 
good enough to keep the heat flow below this level, which dictates thick 
pieces of good insulating qualities on all sides of the enclosure.  This 
probably precludes plexiglass without insulation, although the lower the 
maximum rate of temperature change in the room, the less insulation is 
required.

Another goal is to keep any heat-dissipating electronic components either 
outside the box or located at the top near the heater.

In my force-balance seismometer box, I have about 1-1/2" foam insulation on 
all sides of the inside of a 1/2" plywood box.  At the top are two 
resistors that dissipate about 2 watts total.  I've never tried it without 
the resistors in place.

Karl


--On Tuesday, July 10, 2001 6:06 PM -1000 ian  wrote:

> I'm a little confused about how heating the top of the cover works.
> Won't this just create an inversion layer at the top of the enclosure?
> Ie, instead of the convection currents going all the way to the top and
> turning around, the currents will turn around when they meet the heated
> layer, perhaps an inch or two from the top.
>
> Has any measurable reduction in noise been noticed between the heaters
> being on and off?

__________________________________________________________

Public Seismic Network Mailing List (PSN-L)