o1bigtenor via Dng said on Sat, 19 Oct 2024 11:57:05 -0500


>The process I am working toward controlling will have a water bath for
>process control temperature. I must have the air over the process 1 C
>warmer than the process. Vat T must be provably uniform (constant slow
>speed stirring).

How are you going to prove it? How many temperature sensors will be in
the vat to prove uniformity?


> Inputs from a 80 (or 85) C hot water and 4 C cold
>water will be used to either heat or cool the structure.

Are you removing water to compensate for the hot and cold water you're
putting in? Or am I misinterpreting something?

Process is
>monitored to +/- 0.2 C and I want to hold that T as closely as
>possible to 1 C over recommended T - - - no more and no less (if I can
>achieve +/- 0.5 C in the process i will consider moving to + 0.75 C
>over recommended T. So its not exactly a simple control.

Are you controlling the air temperature, or the temperature of the
liquid in the vat? Or am I completely missing the picture?

Every temperature control system I've seen depended on hysteresis,
which is hard to narrow down to less than a couple degrees Fahrenheit.
I'm wondering if you should take a look at the algorithm from openntpd
to see how they use a continually calculated drift number to prevent
the need for sudden jumps.

SteveT

Steve Litt

http://444domains.com