Hi Ed, i've built many types of temp control
systems over the years, and PID is probably real
overkill and totally un-needed for a system like
this. Most likely you will be using a standard type
mains AC heater, which is essentially an on/off
device. Add to that the huge thermal lag of your
system, ie time from heater turns on to water temp
actually rises, and you are going to get significant
hysteresis from that anyway.

You will get great results from an on/off system
with high gain, you don't need any deliberate
hysteresis because you have enough in the physical
properties of the heater and water.

Just detect the water temperature, using your sensor
mounted very close to the heater and switch the heater
when temp drops below the setpoint. You will probably
get a few seconds thermal lag anyway, but if not try
and switch the heater with at least 5 seconds on,
if your period is too short you risk element failure
from the continual thermal stresses. It should regulate
within 0.1'C with no probs.
-Roman


Ed Edmondson wrote:

> You haven't seen any PID controlled TEC circuits I could use to keep my 1
> gallon Betta fish tank at the same temperature (approximately). I figured I
> would also add a turbidity sensor to run for about an hour or two daily to
> keep the debris on the bottom from collecting and stagnating the water.
>
> I know this sounds like an over kill but experience is a good teacher. I was
> going to measure the water temp and control functions on a 2 X 20 character
> display. The sensor would be an LM34/35/AD590 encapsulated
>
> Thanks for the links and advice. If you chime in please be my guest.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.