Triac to controll relays

[embedded_sys]

In my current design, I have several 12V coil relays that control 120VAC coil relays with a 3 phase 480VAC load. I would like to switch my 12V coil relays out for triacs to reduce cost and save on board space. Can I replace my 12V coil relay with traic like this? I drive the relays with a ULN2003AD. Is there anything special I have to do to switch them out or can I just swap out the relay for the triac? Is there any advantage to keeping the relays over the triacs?

Also, several relays use both the NC and NO sides of the relay. Are there any triacs that can accommodate both NC and NO like a relay does?

 

[AnalogKid]

Can I replace my 12V coil relay with traic like this?

No.

I drive the relays with a ULN2003AD. Is there anything special I have to do to switch them out

Yes.

or can I just swap out the relay for the triac?

No.

Is there any advantage to keeping the relays over the triacs?

One HUGE thing, way beyond a mere advantage. A relay has true galvanic isolation between whatever is driving the coil and whatever is being switched by the contacts. A TRIAC certainly can switch AC - that's what they were invented for. But the TRIAC's gate drive is referenced to MT1, which means the control pin is connected to the AC powerline through some parts that are NOT totally isolating. This is why TRIAC circuits usually have an optoisolator as the input. This is the core of all AC solid state relays, which is essentially what you are describing. You can whip up a circuit from the datasheets and app notes, but it will be dangerous to use. Commercial solid state relays are rigorously inspected and tested, and must meet specific electrical and mechanical design rules to be certified.

Also, several relays use both the NC and NO sides of the relay. Are there any triacs that can accommodate both NC and NO like a relay does?

No. However, there are complex solid state relays with both contact forms.

ak

 

[dknguyen]

Another advanage of relays over triacs: triacs need a big heatsink. Relays don't.

Relays wear down whenever they switch but not while they conduct so use relays where you conduct more than you switch.

Triacs are less efficient than relays but do not wear down when they switch so use triacs for applications where switching is frequent.

 

[MrAl]

In my current design, I have several 12V coil relays that control 120VAC coil relays with a 3 phase 480VAC load. I would like to switch my 12V coil relays out for triacs to reduce cost and save on board space. Can I replace my 12V coil relay with traic like this? I drive the relays with a ULN2003AD. Is there anything special I have to do to switch them out or can I just swap out the relay for the triac? Is there any advantage to keeping the relays over the triacs?

Also, several relays use both the NC and NO sides of the relay. Are there any triacs that can accommodate both NC and NO like a relay does?

View attachment 108233

Hello,

A triac should outlast a relay in a circuit that has to be switched a lot.

A triac turns on when gate current is applied. The difference between an emulated NO or NC relay is the way the circuit is set up. For NO you just drive the gate when you want to turn it on. For NC operation the triac is kept on until you want to turn it off and then the signal is either removed or the gate is shorted out.

 

[dr pepper]

Triacs can usually switch a few 100 watts without a 'sink, relay's shouldnt be a problem, one issue you might have though your triacs are wired to the power line, so either your control circuit will be live or you'll have to use opto isolators or some means of isolation from the triacs to the control circuit.
Some examples:
https://electronics.stackexchange.c...triac-from-a-microcontroller-for-low-voltages
For relays zero crossing switching opto's are good as you wont get much emi.

 

[dknguyen]

Triacs can usually switch a few 100 watts without a 'sink, relay's shouldnt be a problem, one issue you might have though your triacs are wired to the power line, so either your control circuit will be live or you'll have to use opto isolators or some means of isolation from the triacs to the control circuit.
Some examples:
https://electronics.stackexchange.c...triac-from-a-microcontroller-for-low-voltages
For relays zero crossing switching opto's are good as you wont get much emi.

A few hundred seems to be pushing it. Most of the triacs I see are in packages that have thermal resistances of 50-70C/Watt or higher to ambient when not heatsinked. Triacs seem to hang around 1.5V drop at their max rated current so that's 75-105C rise above ambient right there at 1 amp which is about a hundred watts at mains voltage.

 

[dr pepper]

You might be right, I was basing that on a light controller I have that states it'll control 400w loads , and it doesnt have any 'sinks on the triacs.
And I'm in the Uk, our power is 240v so current is less.

 

[dknguyen]

You might be right, I was basing that on a light controller I have that states it'll control 400w loads , and it doesnt have any 'sinks on the triacs.
And I'm in the Uk, our power is 240v so current is less.

Wow I wonder how they do that, even with the 240V, or if it's meant to operate at 400W for any length of time.

 

[embedded_sys]

Sounds like I'll stick to the relays. The 120VAC coil relays pull about 100mA of current when engaged, so that would be about 12 W.

 

[AnalogKid]

Wow I wonder how they do that, even with the 240V, or if it's meant to operate at 400W for any length of time.

A 400 W load at 240 Vac will cause a TRIAC to dissipate fewer than 4 W. The in-wall light dimmers in my house (120 Vac) are rated for 600 W loads. They use the metal faceplate as a heatsink. I've never loaded one to 100%, but at 200 W the front gets noticeably warm, but nothing near dangerous.

ak

 

[dknguyen]

A 400 W load at 240 Vac will cause a TRIAC to dissipate fewer than 4 W. The in-wall light dimmers in my house (120 Vac) are rated for 600 W loads. They use the metal faceplate as a heatsink. I've never loaded one to 100%, but at 200 W the front gets noticeably warm, but nothing near dangerous.

ak

Ah, well if there is a heatsink that's totally different.

 

[dr pepper]

I have not measured forward drop on a triac, but I spose its a diode junction 0.6v, 400w at 240v is about 1.75 amps, 1.75 * 0.6 isnt a great deal of dissipation so just the triac tab might be enough to handle it.
My washing machine has some tiny little smd triacs controlling the water pump & stuff.

 

[ChrisP58]

I have not measured forward drop on a triac, but I spose its a diode junction 0.6v, 400w at 240v is about 1.75 amps, 1.75 * 0.6 isnt a great deal of dissipation so just the triac tab might be enough to handle it.
My washing machine has some tiny little smd triacs controlling the water pump & stuff.

A triac is not a single PN junction, so their forward voltage will be higher than 0.6V. Typically they run about 1.5 to 1.8Volts. But the dissipation is still small compared to the load power they control. And often lower than the power of the relay coil that they replace.

 

[dr pepper]

Ok fair enough, I didnt know that.

 

[schmitt trigger]

Your proposed Triac circuit has some omissions:

1) The MT1 should be referenced to +12vdc, to ensure that there is a path for the current to flow. It will also allow you to use a 3-quadrant Triac.
2) You require a resistor in series with the gate. Calculate based on the Triac's triggering current.
3) Lastly; the 120 volt coils may be highly inductive, which may likely require a snubber in parallel with the Triac.