Really, the whole issue is the tolerance you have in the variation
in temperature. You can always reduce the swings with PWM vs on/off. I
had a problem controlling an oven, and with on/off, we were getting too
much temperature swing, we switched to PWM, and we eliminated the
problem with simple PI control. You might have to experiment and see if
on/off is "good enough". Even with PWM, you can have problems due to
thermal inertia when heating up. By the time you get to the temperature
of the air you need, the metal of the oven is hot enough to keep the
temperature rising, even if you shut off the heat. Well, the same
problem you have with any system that has a large time constant.
We could also argue whether he should use PID or Fuzzy control,
assuming he uses PWM.
Don.
On 3/6/07, Larry
<m("bellsouth.net","leaton10");">
net> wrote:
I agree. I've done this before. You design in a little
hysteresis and
you're all
set. Just turn the heater on a couple degrees below where you turn it
off.
You have plenty of time. You could use a relay, the contacts going to
the heater.
You could use a DIO pin driving the base of a transistor to operate the
relay coil.
You would of course need to have a back EMF diode across the coil as
well.
Larry
Gonzalo A. de la Vega wrote:
> --- In .com
<mailto:ts-7000%2540yahoogroups.com>,
> "Yan Seiner" m("...","yan");"><> wrote:
> >
> > --- In .com <mailto:
ts-7000%40yahoogroups.com>,
> Joel Winarske <joelw@> wrote:
> > >
> > >
> > > > I'm kicking around the idea of controlling an
ordinary 120 VAC
> > > > incandescent light (actually a 150W heater).
> > > >
> > > > I'd like to control the heater based on the
temperature of the
> surface
> > > > beneath it. I need to control it to approximately
105 deg F (more or
> > > > less; great accuracy is not needed here.)
> > > >
> > > > I've found some basic PIC circuits on the web but I
wonder if anyone
> > > > here has any suggestions, both for the temp probe
and for the heater
> > > > control.
> > > >
> > > > As this will most likely be a one-off, an
inexpensive commercial
> > > > solution would be fine.
> > > >
> > >
> > >
> > > You would need:
> > > 1. Zero Crossing input
> > > 2. DC Power circuit
> > > 3. PWM / Triac Control
> > > 4. Load current protect circuit
> > > 5. temp measurement - A/D
> > > 6. Control interface - RS485/CAN/RS232
> > >
> > > The NXP LPC2xxx is a good fit for this.
> > >
> > > How many channels do you need?
> > > How far from the controlling device do dimmer channels
need to be
> > located?
> > >
> >
> > Sorry, I should have been more detailed....
> >
> > I want to run the temp probe into the A/D inputs on a TS7xxx
board,
> > and then have the board control the heater.
> >
> > The reason for this is that I also want the ability to set up
a webcam
> > and multiple control channels, and I want web-based
programmability.
> >
> > So I'm really looking for a 120VAC dimmer that can be run by
a TS7200.
> > All I've been able to find are PIC based designs, which I
could
> > probably adapt, but I am not a hardware person...
> >
> > --Yan
> >
>
> Depending on the thermal conductivity and thermal capacity of what
you
> are controlling, and the precision you need, may only need a
> thyristor: usually the time constants of temperature controlled
> systems is quiet large (like tens of seconds), so just turning the
> heater on and off is enough. If this is the case you won't need a
PWM
> output either, just binary.
> Note that by saying a thysritor I mean any on-off device: triac, or
> relay too. A thyristor is rectifying, so you'll get at most 1/2 the
> power unless you use two of them.
>
> Gonzalo
>
>
--
Dr. Don W. Carr
J. G. Montenegro 2258
Guadalajara, Mexico
+52-333-630-
0704
+52-333-836-4500 ext 2930