>
> Terry Dagnin wrote:
> >
> > hi, i am interested in a project for controlling inductive
> > (mains transformer type ) loads using a pic. problem is the actual
> > controlling of the load. i believe a thyrister is not a very good
> > idea. what is better ?

What do you mean by 'control'?  If you mean phase control of power
levels or voltage (as in a light dimmer) then you should be able to use
triacs.  However, the problem is complicated by the fact that a
transformer's primary impedance is actually composed of resistance,
leakage inductance, self-inductance and a reflection of the load
impedance of the secondary.  You cannot rely on it being purely
inductive or resistive.  Also, transformers require a zero volt-
second product (averaged over time) to avoid saturation - see
below.

Thyristors require a zero-current period to turn off.  If the
zero-crossing is caused by reversal of the load current because
of the AC supply, then this is termed 'naturally commutating'.
Circuits can be designed to force a thyristor to turn off (i.e.
Reverse Impressed Voltage or Current Impulse Displacement) but
it is simpler to use Gate Turn Off thyristors which are specially
designed so they can be turned off by pulsing a special gate
pin.  Such 'forced commutation' is only required if there is
no natural zero-current period.

Assuming you have a low-frequency (i.e. below a few hundred Hz)
AC supply, thyristors are a good choice for power control.  When
driving inductive loads, the device will not turn off at zero
voltage crossings of the mains supply, because the inductance keeps
current flowing at these points.  You will need to carefuly
analyse the circuit at all conditions for predictable performance.

The non-sinusoidal current waveforms induce noise on the AC supply
line (and the load).  The power factor (real divided by apparent
power) can cause unexpected stresses on components and the mains
supply.

Driving a thyristor at mains potential requires the entire
driving circuit to be floating at mains potential, or some form
of optical or transformer isolation to be employed.  Optical
triac drivers are very cost effective.  For proper control, you
will need to know at least the zero crossing points of the mains
voltage.  Preferably, you should also have a current sense so that
the thyristor turn-off point can be known.  Any such sensors will
also need optical or other isolation.

I have noticed that switching a transformer using a triac increased
the audible hum of the transformer, even though the triac was
continuously biased 'on'.  My theory as to why this happens is
because of a slight asymmetry in the triac's forward and reverse
conduction.  This causes the magnetic flux in the transformer to
become biased in one direction.  At one peak of the mains this
causes saturation and a consequent large current spike in the
supply.  Perhaps this is not the correct reason, but it is certainly
important to ensure the 'volt seconds' of an inductor or transformer
averages to zero.

Regards,
SJH
Canberra, Australia