Go Down

Topic: why does an inductive spike trigger a triac? (Read 2979 times) previous topic - next topic

seanz2003

Sep 08, 2013, 12:57 am Last Edit: Sep 08, 2013, 02:54 am by seanz2003 Reason: 1
I am using a triac (http://www.nxp.com/documents/data_sheet/MAC97A8_A6.pdf) to switch a small 120volt 5VA solenoid. The circuit utilizes a MOC3041 optoisolator and is similar to Figure 13 from the datasheet: http://www.taydaelectronics.com/datasheets/A-036.pdf. The main difference with my circuit is that I had to remove the 330R (between pin 8 6 and "HOT") because it wouldn't trigger with it in place (because its 120v not 240v like in the datasheet???). The circuit seems to work fine except that when the clothes washing machine shuts off, the resulting spike on the 120v mains line briefly triggers the triac. Can someone explain why? Now the simple solution is obviously to pick a different 120v circuit to power my work bench, and leave the washer/dryer to their own devices, but i would rather take this as a challenge to improve my design.  Besides the snubber shown in Figure 13, what other things can I do to prevent transients on the mains from triggering the triac?

runaway_pancake

#1
Sep 08, 2013, 02:06 am Last Edit: Sep 08, 2013, 02:10 am by Runaway Pancake Reason: 1
Do you have MT2 connected to Line? And the Load connected between MT1 and Neutral?


The circuit utilizes a MOC3041 optoisolator and is similar to Figure 13 from the datasheet: ...

What's wrong with Fig. 12?


The main difference with my circuit is that I had to remove the 330R (between pin 8 and "HOT") because it wouldn't trigger with it in place (because its 120v not 240v like in the datasheet???).


There's no "pin 8", it's pin 6

> > > Adding Dwg.
"Who is like unto the beast? who is able to make war with him?"
When all else fails, check your wiring!

seanz2003

Quote
Do you have MT2 connected to Line? And the Load connected between MT1 and Neutral?
No I have it Mt1 (pin3) to Line and Mt2 (pin1)to load-> nuetral. I matched the schematic for the opto to the pinout of the triac. Is it backwards? though I do believe I tried it both ways.
Quote
What's wrong with Fig. 12?

3041 is specifically listed on figure 13, so I thought it would be the best bet.  Is that a more appropriate way to do it?
Quote
it's pin 6
Right. typo. its a six pin package. corrected
Thanks for your drawing!

runaway_pancake


No I have it Mt1 (pin3) to Line and Mt2 (pin1)to load-> nuetral. I matched the schematic for the opto to the pinout of the triac. Is it backwards? though I do believe I tried it both ways.

No, MT2 and MT1 are very different, not interchangeable.


Is that a more appropriate way to do it?

Yes, I believe so.

My circuit is for a random-fire driver, it may benefit from the damping effect of the 2k4 w/ 0u01.

Use HV rated, hi-W components.
"Who is like unto the beast? who is able to make war with him?"
When all else fails, check your wiring!

dc42

#4
Sep 08, 2013, 04:49 pm Last Edit: Sep 08, 2013, 04:53 pm by dc42 Reason: 1
Your circuit looks ok to me. The washing machine is probably producing a transient that exceeds the dV/dt rating of the triac. Connecting a resistor of about 100 ohms to 1k between G and A1 of the triac may fix it. Too high a resistor will have no effect and too low a value will prevent the triac being triggered by the opto coupler. If that doesn't work then you will need an inductor in series with Line.

PS if those component values were intended for 240V, then for 120V you may need to reduce the 2k4 resistor to 1k2.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

drjiohnsmith

are you missing a resistance from pin 4 of the opto to the triac ?

seanz2003

Quote
are you missing a resistance from pin 4 of the opto to the triac ?
Yes, but i do have resistors on pin 6.
Quote
Do you have MT2 connected to Line? And the Load connected between MT1 and Neutral?
This is actually how I had it connected; my recollection was wrong.
Your circuit seems to work much better. (I used 1k and 330 resistors). It still fires about 1 out of 15 times, but that's a great improvement! From a circuit analysis perspective, can you explain whats going on? I'm guessing the 0u01 capacitor is absorbing the voltage spike?

MarkT

Quote
why does an inductive spike trigger a triac?


The literal answer is probably MT2-gate junction capacitance - spike on the MT2 couples to the gate via
the intrinsic capacitance and forces enough current to flow in it to trigger conduction -
also known as dV/dt triggering since I = C dV/dt.
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]

runaway_pancake


Your circuit seems to work much better. (I used 1k and 330 resistors). It still fires about 1 out of 15 times, but that's a great improvement! From a circuit analysis perspective, can you explain whats going on? I'm guessing the 0u01 capacitor is absorbing the voltage spike?


With a 1K (instead of 2400) you're changing the time constant. 
Use two 1K and a 330 in series (=2330), that should be close enough.
[If you don't have another 330, then try with the two 1K.]

Yes, look at it as filtering the optotriac (SBS) output.
"Who is like unto the beast? who is able to make war with him?"
When all else fails, check your wiring!

drjiohnsmith

I still think u need that gate bleed resistor, as per the app note,

Go Up