On Sat, Apr 06, 2002 at 04:21:26PM +0530, Anand Dhuru wrote:
> Hi folks,
>
> I have built a circuit around a 16F628 that drives a Triac directly (PIC - 220 ohms - gate) and it works fine with incandescent lamps / bulbs. Since I am not interested in controlling the phase angle as such (the load is to be on, or off, the intensity is not to be controlled) I am not pulsing the gate of the triac; the pin driving the gate is always either on, or off.
>
> Now, my question is, can this setup be used to switch heavier loads such as
>an air-conditioner, if I use a triac of an appropriate rating?

There are multiple immediate issues. First is that the AC has inductive
elements.  The second as you mentioned is current rating. The third is that at
that load will cause significant power loss across the triac. Also there's
a good chance you'll blow the PIC because there's no isolation between line
level and the PIC. Usually in this circumstance I'll use a optoisolated
triac drive like a MOC3010 to isolate the control circuitry from line level.

Many moons ago I asked a very similar question in the DC arena about driving
DC motors for a robot. Roman Black, who I just realized that we haven't heard
from in the last month or so, helped tremendously. Also Fr. Thomas McGahee
contribited a thought or two. First your questions then their thoughts.


>
> Which brings me to the second question! Even triacs rated at 12 Amps, 400 volts come in a TO220 package. Can these devices really control the load of an air-conditioner without blowing up? Would a heat sink be required?
>

It won't work without a snubber network to help manage the inductive load.
A heat sink would definitely be required.

> Incidentally, I have 230 volts AC in my country, and the AC I'm interested in controlling would be 1.5 tons.
>
> Any advice here would be greatly appreciated, folks!

Here's the advise I received:

1) Use a relay. Actually at this load you'd probably need a contactor. Reasons
are pretty simple. You're not trying to pulse the load. The relay will be
minimally affected by the inductive load, and the power loss in operation
is negligible, and gives control circuit isolaton..  Thanks Roman,
wherever you are.

2) A hybrid system can assist in the one area where relays have issues, making
and breaking contact. When a relay makes contact, there's often arcing because
the switchover isn't done at the zero crossing. This damages the contacts
of the relay over time. Fr. McGahee described a system where a triac is
used in conjunction with a relay to help save the contacts from arching.
Basically you wire the triac and relay in parallel. So switch the load on you
first switch the triac, which turns on the load without arcing. Once the
load comes up to speed kick in the relay and drop out the triac. The relay
then carries the load. At dropout do the reverse. I actually found that thread
in the archives. I cannot believe that it was THREE years ago! Here it is
(NOTE!!! PASSWORDED SITE. Both login and password is 'piclist'):

http://www.infosite.com/~jkeyzer/piclist/1999/May/0958.html

A read of that thread can give you some other thoughts on this exact same
subject.

Good Luck. And use a relay.

BAJ

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu