Triacs - interesting YouTube vid

[android]

I just stumbled across this and found it a quite interesting technical explanation of how triacs work. It may be interesting to others (I have no affiliation...it caught my eye enough to subscribe though):

https://youtu.be/I69Fiy3IwBE

Enjoy

[T3sl4co1l]

Nice!  Basics like this -- what parameters are and how they are measured -- are invaluable!

Tim

[wilfred]

Thanks. I suppose this is how I am going to find interesting videos in future when YT has everyone fighting for thumbnail attention.

[BravoV]

To me, the quality for electronics educational video, this is top notch, thanks. 

[rhodges]

I really enjoyed watching these three videos. They explained a lot of the fuzzy ideas I had about triacs, and I am getting motivated to do some playing with them.

At the very end, he calculated the gate resistor value to keep the gate current under the maximum (2 amps). I note that the gate current falls very quickly (one or two microseconds?) when switching on. But my question is, "do I want to be hitting the gate that hard?" when only ten milliamps (for his triac) would be enough. Did he choose the lowest reasonable resistor value to ensure current at the ends of the phases (eg, near crossover?) Is there another reason?  At mains voltage, that could be a lot of power for those microseconds.

[David Hess]

But my question is, "do I want to be hitting the gate that hard?" when only ten milliamps (for his triac) would be enough.

If the trigger current is low, then the delay and turn-on time are longer leading to a situation similar to that of a bipolar transistor in secondary breakdown which can damage the thyristor over time.  So it is better to have a hard turn-on than a soft turn-on.

[emece67]

.

[T3sl4co1l]

Yeah, typically you'll drive the gate with a modest current, say 100mA.  A number of the parameters are measured at this condition, too.

Or use a typical driver, like for an AC switch, an MOCx optotriac and a 220 ohm series resistor, or thereabouts.  The phototriac's peak current rating is about the same as the gate, but it only delivers current while there's some MT1-2 voltage, so it's pulsed only very briefly.

Don't think I thought about the on-state gate voltage before, but it makes perfect sense.  It's always an important question -- you must never fall into the trap of thinking a device is unidirectional only, or nonreciprocal.  Many of our devices are, or can be designed according to such principles (e.g., analog and digital logic -- opamps and gates), but they are complicated devices, made specifically to satisfy that constraint.  The building blocks they are made from, are more nuanced, and it helps to be aware of these things.

I was aware, however, that GTO (gate-turn-off) SCRs are a thing, and that you yank current out of the gate (mostly equal to I_F!), probably with a bit of negative supply to facilitate it, but mostly you're pulling it down from a diode drop, and it's technically sufficient to just short that out (at least, if series resistance isn't a problem).

So you'll get the same thing in a TRIAC, of course.  Probably can't GTO one, though!

BJTs also do this a bit: at low Vce (in saturation), the B-C junction forward-biases and Vbe drops slightly.  This can be used as saturation detection, and at least one designer used this mechanism for a relaxation oscillator, which as a result generated a very accurate amplitude (because it's pulling down to very nearly 0V).

Tim