by Tony » Tue Oct 14, 2008 2:06 pm
The basic idea is certainly sound, but there are a few pitfalls for the unwary when experimenting with these types of motor control circuits.
Triacs can sometimes have great difficulty turning off at the zero crossings with highly inductive loads. This can cause erratic operation at certain phase angles and motor loads. By far the better way is to use back to back SCRs rather than a Triac, if at all possible. It will be far less temperamental.
The IGBT circuit is even better still, but it has one significant vulnerability to failure.
The mains supply voltage often carries very fast high voltage spikes. These originate throughout the whole power distribution network, and especially from lightning strikes inducing voltages into overhead transmission lines.
These voltage spikes can easily reach several thousand volts, but only for microseconds, and do not effect most mains operated equipment. Now something like a Triac or SCR will just turn hard on for half a cycle through dv/dt switching of the gate. In effect, in turning on, it protects itself from high voltages. But an IGBT will not do this. A 600 volt rated device will be destroyed by the very first mains borne voltage spike significantly over 600 volts.
If using an IGBT, make damned sure you have transient voltage suppression, preferably right across the IGBT itself. And maybe a second suppressor right across the incoming mains supply.
All the circuits in that application note are directly connected to the mains supply which can be pretty lethal. I much prefer to isolate the drive to the power control device, either with an opto isolator, or pulse transformer.
The microprocessor, or any control potentiometer is then safely isolated from the mains voltage. It also means you can use test equipment to monitor voltages and waveforms, without killing either the test equipment or yourself.
Also known as the infamous "Warpspeed" on some other Forums.