Triac gate turn on time

[fuseshut]

I currently turn a triac on using an optocoupler MOC3052. I saw when writing the code that I need to pulse the MOC3052 optocoupler(from the microcontroller digital output)for a pulse width of 400-500us. I tried smaller pulse widths and couldn't get the triac to come on every zero crossing. Any one got any reference on maximum turn on gate time alotted for a triac. I am wondering if I am killing the triac by giving it a pulse of longer than 4-10us. Doesn't seem like much when I am trying to pulse within every 1/2 cycle (every 8.3ms).

 

[itsmoked]

I'm surprised you get that short a time ON to work. It could be you are scorching them because of the short time. You should be turning them ON and leaving them ON until you want them off. Not just pulsing them. A really short low energy turn on pulse causes the ignition current to start at one little spot on the structure and slowly spread out. This is a bad thing as the conducted current then conducts only thru that tiny area for the first bit of time overheating it badly. That can damage the part. You need a good bit of turn ON current to keep them happy. Look on the data sheet and see what the minimum turn ON current is and make sure you exceed it handily.

Also turning off inductive loads will kill them if you don't handle them correctly.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

Not quite, Smoked. The destruction of the triac can not be avoided by applying the firing pulse for a long time. And it is seldom a problem with triacs. More so with thyristors where the silicon tablet is quite large and it takes some time for the conducting area to cover the whole chip. One uses inductive components (reactors) to limit the rate-of-rise of current, usually to values like 50 or 100 A/us so that the current density doesn't get too high.

The reason a triac doesn't fire is probably that the gate pulse is so short that the load current doesn't reach holding current. This is never a problem with resistive loads like lamps and heaters, but can be quite tricky when the load is a motor winding or other coil like a transformer primary. Putting a resistive load across the winding helps. It is also a good way to diagnose the problem. If you put a 25 or 40 W incandescent lamp there and triggering gets OK, then you know that the problem is getting up to holding current before gate pulse is removed. Als, most triacs need more gate current in 3rd quadrant than in 1st. And more gate current if cold.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

Gunnar you're saying the reactor is in the power circuit?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

Yes!

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Skogsgurra]

Like this:

http://www.eng.uwi.tt/depts/elec/staff/rdefour/ee33d/s4_prtcn.html

Took some time to find that link. You owe me a beer! (I know, I can have yours, too )

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

Only! if the page actually works... It's not at the moment.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

What about this then?

http://www.pwrx.com/pwrx/app/solid_st.pdf

Section 22 is about di/dt.

That first link wouldn't open for me either. And still, I have a copy that I got early trhis morning. Try later.

Do I get the beer?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Skogsgurra]

..in chapter 5..

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[fuseshut]

Gunnar,
are you saying that by pulsing the triac (opto) with a pw of 500us shouldn't kill the triac? THe load is a 500W, 120Vac incandescent bulb. SO for sure I am getting over the holding current. And I have verified on a scope looking at the triac trigger pulse width versus my zero cross detects, that everthing looked good. I am wondering if my circuit operates at 130VAC, with a 178ohm resistor in series with the high side of the optotriac and gate of the Littlefuse triac, if that's killing the triac. Maximum surge current is 1A on the littlefuse. @ 130vac, peak is 183, which gives me a current over a 1A. What blows my mind is that I have runned this circuit in the lab, @ full load, 500W, over a period of 200+ hours, with the microcontroller pulsing the opto with PW of 500us of trigger to the triac, operating as a light dimmer, dimming and have never seen a failure. I am able to change my analog control to the circuit and watch the dimming behave beautifully. Yes the triac does heatup to 130degF, but it's stable and the dimming still behaves with a 500W load. Boards are sent out in the field, and they are failing.

 

[Skogsgurra]

It would help a lot if you gave all pertinent information. It is only now that we learn that you have a 500 W bulb connected. What else do we not know? Rated current of triac? What gate current are you feeding into it?

If you keep gate current within rating, you can never kill a triac with a long trigger pulse. No way.

It may actually be a di/dt problem. At that rather high load the di/dt can be too much for the triac, especially in the third quadrant and at 50 % output (highest voltage, highest di/dt).

Read the data sheet and see if your RMS is within specs. Also find out what gate current is needed in the third quadrant and make sure you have enough. Also, make sure you do not put too much current into the gate. Keep specifications.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

Thanks Gunnar for the tome.. I shall read the designated chapter.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[fuseshut]

Sorry for skipping right to the climax:
Load = incandescent bulb, 500W
voltage is 120VAC
triac is standard (designed for light dimmer), 8A rated, 600V
there is a snubber (0.1uF cap with 100ohmr resistor in series) across triac terminals
also there's an MOV across terminals

 

[itsmoked]

Cold tungsten is 1/17 the resistance of the operating temperature resistance.


P=V²/R

R=V²/P
R=120²/500W
R=29 Ohms (hot)

Rcold = 29/17 = 1.7 Ohms

120/1.7 = 70A

It will likely be a little less because of inductances and wire sizes but you will occasionally be hitting something around there. Got that covered in the device spec?






Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

Yes, Smoked. That is probably it. A 500 W lamp turned on at peak voltage (50 %) is tough. It even eats mechanical relays at a rate you wouldn't believe.

Did you do your in-house testing with a lamp or a heater? There's a bigg difference in initial current.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[fuseshut]

Oh yes, had a 500w underwater light placed in a bucket of ice water for days while circuit was running, dimming from 0-100% brightness with no problem. The datasheets describes these "standard" triac with built in snubbers to handle phase control/light dimmer applications. I also have an external snubber across the triac terminals.