Triac Dimmer not turning off during negative going half cycle of AC main

[Nazmul Hasan]

Posted on February 05, 2018 at 14:13

Hello,

I got a problem with my light dimmer pet project using ST's snubberless triac, T435-800T. It's working fine (turning on and off based on the control input) with the positive going half cycle of AC main. But during negative going half cycle, I found a strange behavior. Traic failed to turn off  during the negative going half cycle of AC main. I tried two separate rated incandescent lamp as a load. One rated at 25 W and another 60 W. What I found are, for 25 W lamp, turning off of Triac during negative half cycle not working at random intervals (please find the attachment) and for 60 W lamp, there is no turning off of Triac during negative going pulse. Please be noted that I checked control signal for both cased, and controller producing exact signal as the positive going half cycle.

Can you please share your experience if I got same problem earlier? Do I need a Triac with higher Irms (on-state) value to overcome the issue? T435-800T has a Irms (on-state) of 4A.

Thanking you.

renard.benoit

#dimmer #t435-800t #triac #snubberless

[Benoit RENARD · Accepted solution]

Posted on February 05, 2018 at 14:48

Hello Nazmul,

It seems that the Triac is staying ON if the di/dt at Line zero current ((di/dt)off) is higher than the Triac maximum (dI/dt)c (Triac turn-off capability). But at 60 W / 230 V / 50 Hz, the (di/dt)off due to the load is approx. 0.13 A/ms when the device maximum (dI/dt)c is 2.5 A/ms @Tj = 125°C. Then ok (Tj should not be higher with this low wattage load).

Regarding the control signal, how long is apply the gate current to the Triac ? Is the gate current removed a long time before the load zero current ?

Is the Triac integrity OK ? Same problem with a new one ?

Note: According to schematic, the applied gate current is 12 V / 180 Ohm = 6.7 mA. Too low. Must be egal or higher than Triac IGT, i.e. 35 mA (no link to the issue mentionned here).

Let me know

Best Regards

Benoit

[Nazmul Hasan]

Posted on February 06, 2018 at 12:38

Hello Benoit,

Thank you very much. The main reason was the width of gate pulse. I tried with not to cease the gate pulse enough before the zero-cross. Now, I'm providing a pulse of single ms at every zero-cross and working fine. I tried with lower width of 250 us but sometimes it failed to turn on the Triac during positive going half cycle of AC mains. Can you please let me know the safest width of gate pulse to trigger Triac at QII and QIII (say for ST's Triac)?

Can you please let me know the the formula to calculate dI/dt(off), which you calculated 0.13 A/ms for 60W load. Any recommended ANs on usage of Triac in practical way?

Regarding you note, I'm afraid you multiplied with 100 (instead of 1000) to calculate IGT in mA.

Regards

-Nazmul

[Benoit RENARD]

Posted on February 15, 2018 at 11:32

Hello Nazmul

And sorry for the delayed reply.

250 us is may be too short regarding the latching current (IL) of the device. You have to apply the gate current until the load current is overpassing the device IL (IL max = 50 mA in Q2 and 60 mA in Q3 according to the T435-600B datasheet).

Time to apply gate current versus IL is defined in this application note: AN303 : Latching current.

Regarding the (dI/dt)off, I have made a rought calculation :

- sinusoidale RMS current is IRMS = 60 W / 230 V = 260 mA.

- (dI/dt)off @ zero crossing voltage (derivative of 50 Hz sinus @t=0) is approx IRMS / 2 then 0.13 A/ms.

Let me know if you need more

Best Regards

Benoit

[Nazmul Hasan]

Posted on February 17, 2018 at 09:01

Hello Benoit,

Thank you very much for the explanation and the reference of latching current AN. Now, it made much more sense to me. I will definitely reach to you again but please let me fail first.

Regards

Nazmul

[Benoit RENARD]

Posted on February 19, 2018 at 09:46

Hello Nazmul,

Please dont hesitate if you need.

Testing to obtain experience in Triac driving is the best !!

Regards

Benoit