Hi Andy, Marv,
> A Triac will exhibit a considerable loss during
> normal, post startup, operation.
That loss might be between 1 and 2%, depending on what line voltage is
used (worse for 110V than for 220V, of course). Depending on the
situation, that might be acceptable, or not.
> I've seen circuits
> where a Triac is used with a relay connected in
> parallel. The Triac is first enabled, then shorted by
> the relay. This greatly reduces contact wear on the
> relay while avoiding the typical 2V drop of a Triac.
Exactly. During my 18 years repairing industrial equipment, I also have
come across many such circuits. They make the best of available parts,
combining the controllable TRIAC with the low drop relay. Using this
system, typically soft starters for motors of up to twenty kilowatts or
so are implemented in a small, circuit-braker-sized unit that doesn't
even need any heatsink.
For a 1.5kW amp, it might be acceptable to skip the relay. After all,
anyway we have at least 50% loss in them, so an additional 1 or 2% is
not too terrible. On the other hand, a crude plain old soft starter
using a power resistor and a relay is cheaper, and has lower loss! It's
> Out of curiosity, how does a zero crossing solid state relay work? I had
> thought it had a triac inside, but your comment about voltage drop makes
> me question that assumption.
Yes, normally they use either a TRIAC, or two antiparallel SCRs. And
they do have a significant voltage drop. These thyristors by itself will
always switch off on the zero crossing (of current). True zero crossing
solid state relays have some circuitry that produces a window for the
firing pulse that falls into a small time span just after zero crossing,
so that switch-on is on the voltage zero crossing and switch-off of
course remains on the current zero crossing.
Solid state relays with DC output use MOSFETs, or IGBTs. I can imagine
that there are also AC solid state relays using back-to-back MOSFETs or
IGBTs. These would have lower voltage drop, at least at low current.
Many people ask themselves how you can drive a MOSFET or IGBT without a
galvanic connection, given that they need some sort of drive both for
switch-on AND for switch-off. The answer is simple: An RF oscillator ,
with a secondary coil fitted with a diode and capacitor, allows to
transfer the drive power from the control circuit to the power device.
When using two MOSFETs or IGBTs, it's a simple matter to use two such
secondary circuits. Years ago I designed and built such modules to
control MOSFETs in a switching power amplifier intended to drive DC
servomotors in the 10kW power class. It was funny because all my
colleagues viewed that RF oscillator as black magic. They weren't hams!
Visit my hobby homepage!
Amps mailing list