ATM Build/Design of a Kiln.

[Aplanatic@aol.com]

Sam Michael wrote:

> I was thinking that motorized variac that could increase and decrease
>  voltage at a predetermined rate would be a much better way of controlling
>  the annealing process.  A curve could be plotted showing the heat loss at
>  various gradients of temperature and the exact temperature would be less of
>  an issue so long as your started above the annealing temperature and knew
>  where you were going.  This to me would seem to be a better way of
>  controlling temperature and temperature change.  The elements would remain
>  on all the time, but at a lower voltage.  Once you are adding less energy
>  than the loss of the kiln, the temp would drop.

I think a better way to accomplish fine temperature control is to fully turn 
on and off the heating elements, but do it rapidly enough so that the 
inherent thermal time constants of heater/oven/glass smooth out the 
temperature to the degree desired.  This is basically pulse width modulation 
of the heater.  If you cycle the heater on and off every few seconds and the 
thermal time constant of the glass/oven system is an hour, then this natural 
low pass filter does the work of a more elaborate system using variacs.  To 
control the amount of power, just vary the duty factor.

Why is this better?  Well, it can be done with relays and a bit of software 
instead of the huge and expensive variacs needed to handle this much power.  
In addition, it's very easy to control the state of a relay, whereas a variac 
control scheme requires some interesting hardware, etc.

Regarding the heater windings, they can be used in series/parallel 
combinations to get the best coverage inside the oven while maintaining the 
right overall resistance and power rating. 

To calibrate the thermocouple, most people use small cones that melt at a 
specific temperature.  You observe the cones while heating the oven and when 
they slump you know that you are within a few degrees C of their stated 
temperature.  Several thermocouples should probably be used inside the kiln 
to explore thermal gradients.

The most interesting problem to me is designing the oven, heaters, and glass 
support structures to reduce thermal gradients in the glass as much as 
possible.  If anyone has any comments on this I would like to hear them.

Dave Rowe.