I am making a phase controller to adjust speed of an AC motor. I use BT139-600E triac to switch AC voltage and MOC3052M optocoupler without zero-cross to turn triac on. It works fine, even without noise back to AC line, because I added two LC-filters.

Anyway, I see oscillogram is not so good. As I see, it takes about 800µs to turn triac off. That's why it won't turn off at zero-cross, but about at ±80V. Is that because of the snubber RC-circuit?

So the question is: Is there a way to make it turn off faster? Or maybe you can advise other specific triacs for this task? To play with snubber circuit?

I use the usual scheme to control triac with snubber RC and varistor.

enter image description here enter image description here enter image description here

  • 5
    \$\begingroup\$ The TRIAC doesn't turn off at the voltage zero-crossing, it turns off at the current zero crossing. You need to look at the current waveform in order to understand what's going on. The complex impedances of your filter and your motor are causing a phase shift between voltage and current. \$\endgroup\$
    – Dave Tweed
    Commented Nov 8, 2016 at 12:58
  • \$\begingroup\$ What exactly do we see on the traces, where did you measure things? It certainly isn't a mains voltage waveform... \$\endgroup\$
    – PlasmaHH
    Commented Nov 8, 2016 at 12:59
  • \$\begingroup\$ @PlasmaHH I measure at leg #2 of P2 connector. \$\endgroup\$
    – Bulkin
    Commented Nov 8, 2016 at 13:17
  • \$\begingroup\$ @DaveTweed, I use LC-filter before triac. It couse phase shift between voltage and current, am I right? \$\endgroup\$
    – Bulkin
    Commented Nov 8, 2016 at 13:25

2 Answers 2


This is normal for a Motor drive with low load.

The low Power Factor is due to the motor inductance when little "work" is being done. Under load the current phase shifts towards the voltage the "real" component of energy [W] vs the "Stored" phase shifted " component of energy [VAR]

Meanwhile this reactive energy or excitation current in the motor core also creates back EMF to reduce the current in the windings ( inductive impedance increases with speed) but a current lags almost 90 deg when running with no load.

Since the Triac shuts off near zero current ( I_hold threshold) this explains why you get what you see.

This only problem you may see with the transition power during switching with V*I across triac can cause self-heating during start surge currents of 800 % and abrupt speed changes.

I hope I am explainin' in terms you understand.

Try a current shunt or CT and view V,I(t) or V vs I in XY or product of V*I for a better understanding about losses in Triac.

For a 1st order under approximation, I measure DCR of motor coil and lookup triac ESR by Vdrop @ rated Current and compare. If ESR of Triac / DCR motor > 2% , I would expect a heat sink is needed and if >10% the Triac may be undersized. This is because conduction losses are proportional to R (I^2*R)


You can use a 3-quadrant triac optimized for turn-off with an inductive load. Some makers call these devices 'alternistors'.

The makers claim similar performance to two back-to-back SCRs, meaning that the one quadrant does not affect the commutation in another.

  • 1
    \$\begingroup\$ Thank you, I read the docs and will try another BTA16-600BWRG snubberless triac wich has larger sensivity. Current triac has 5-22mA Igate, but BTA16-xxxxW has 50mA. So it must turn off earlier. \$\endgroup\$
    – Bulkin
    Commented Nov 9, 2016 at 6:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.