TRIACs

[spyro2]

Hi folks,

My questions are about TRIACs.

I'm attempting to drive one from an optotriac.

There are a number of variables, and considerable searching of the net has not come up with much of use.

Q1) Zero crossing - I want to be kind to my load, so zero crossing detection seems like a good idea. Should I select both an optotriac AND a companion triac with ZC, or is it sufficient to  use ZC on just one device? If so, which one?

Q2) Snubber networks - one of my loads will be a small pump (central heating circulator type). It will have some inductance, and I dont want my triac to latch on as a result -  how does one calculate values for the RC snubber network? or is it just "one size fits (mostly) all? Other loads might be more or less inductive - its not quite known what they will be yet. Nothing "agriculturally huge".

Q3) Safety - an "off" mains relay is "kind of" safe - its physically no longer connected. how well do triacs shut off? i'm not planning on touching the terminals on a regular basis, but do they shut of well enough that should one make a mistake, one wont get a nasty surprise?

Thanks for any help!

[David Hess]

Q1) Zero crossing - I want to be kind to my load, so zero crossing detection seems like a good idea. Should I select both an optotriac AND a companion triac with ZC, or is it sufficient to  use ZC on just one device? If so, which one?

Huh?  The optocoupler is the part which includes the zero crossing capability.  The TRIAC is just a TRIAC.

Q2) Snubber networks - one of my loads will be a small pump (central heating circulator type). It will have some inductance, and I dont want my triac to latch on as a result -  how does one calculate values for the RC snubber network? or is it just "one size fits (mostly) all? Other loads might be more or less inductive - its not quite known what they will be yet. Nothing "agriculturally huge"

A snubber network is a good idea no matter what the load is.

https://www.onsemi.com/pub/Collateral/AN1048-D.PDF

Q3) Safety - an "off" mains relay is "kind of" safe - its physically no longer connected. how well do triacs shut off? i'm not planning on touching the terminals on a regular basis, but do they shut of well enough that should one make a mistake, one wont get a nasty surprise?

TRIACs shut off just fine but leakage through the TRIAC's snubber network means the circuit is never really off.

[dmills]

You can still get one hell of a belt from a circuit switched off by a SSR, so I would never rely on a triac for isolation. Apart from anything else they turn on if dV/dt becomes too large (Like say during a spike).

In addition a single pole switch or relay in the off position is not by definition isolation, you should still be treating the circuit as live as you have only cut the phase conductor but not the neutral (Both of which are considered to be live).

Regards, Dan.

[Zero999]

You can still get one hell of a belt from a circuit switched off by a SSR, so I would never rely on a triac for isolation. Apart from anything else they turn on if dV/dt becomes too large (Like say during a spike).

In addition a single pole switch or relay in the off position is not by definition isolation, you should still be treating the circuit as live as you have only cut the phase conductor but not the neutral (Both of which are considered to be live).

Regards, Dan.

Yes that's true. A switched off TRIAC should never be relied up on for isolation and protection against electric shock. Another thing to bear in mind is, because TRIACs can be triggered by spikes and are prone to failing short circuit,  they should also not be used where reconnection of the supply could present a risk of injury or fire, unless other precautions are taken.  For example in a circuit controlling a motor, which could chop someone's hand off, if the motor turns on at the wrong time,, another form of protection is required: i.e. safety relay and contactor. Another example is a heater controller, which could result in fire, if the element is left connected continuously: a thermal fuse should be in place, in case the TRIAC fails short circuit or the controller dies, with its output high.

[David Hess]

If leakage through the snubber and TRIAC is a problem, then I think the snubber can be moved and placed across the load but this requires access to hot and neutral at the TRIAC which is not always available.  Or a power resistor can be placed across the output to shunt any leakage.

As others posted, controlling the leakage should be relied on for safety.

[spyro2]

Ok, so no real surprises there - Triacs aren't any good for isolation.

Thanks for the link to the app note - thats one of the best I've seen on the topic so far for readability. its a LOT to process though.