TRIAC based on/off circuit showing flash problem with LED lamp

[getfast_kiran]

I had made a circuit to turn load on and off using a TRIAC.
It all worked fine until the load was resistive or inductive.

As soon as I connected a LED lamp to the circuit all problems started coming in to the circuit. When the TRIAC was off. The LED lamp started to flash every few seconds with each flash becoming
brighter than the one before it.

So when I read about it in eevblog forum. I found a post where Mr. SeanB had said
" LED lamp with a TRIAC can cause issues of flashing when off, as the slight leakage current through the snubber capacitors can be enough to charge the power supply in the lamp up to startup, which will make the lamp either flicker when off, or flash every few seconds."

But I could not find a solution to it being supplied in the post or across the forum that I could find.

So if you guys could help me with the same. I really appreciate it and thankful for your time.

Here is the modification I did from the schematic. I used the earth as the return for voltage resistor circuit for voltage sensing even though I knew the earth is such a bouncy return path.
Also I removed one of those 100nf capacitors in the 5V stability correction.

Also instead of the 360 ohm resistor and 470 setup changed it to the MOC 3021 datasheet recommended 180 ohm and 2.4k combo resistance along with the 0.1uf X rated capacitor.

Looking forward for you instructions to recitify the issue.

[getfast_kiran]

Following is the circuit used for the opto coupler 3021

[Seekonk]

re you talking about a  you build it LED or a commercial LED LAMP.  I use <19W E27 lamps as night lights sometimes by putting a .1uf or smaller in series with the lamp.  Depending on the electronics inside, I get a nice low power lamp, it flashes, or doesn't turn on at all. All in series triac circuits have leakage issues. A good size RC (100ohm .47uF) across the lamp may solve some issues.

[BrianHG]

As a test, can you try the circuit without the snubber cap?
Without the opto & add a resistive load on the gate to ensure it's off...

[getfast_kiran]

As a test, can you try the circuit without the snubber cap?
Without the opto & add a resistive load on the gate to ensure it's off...

As you told I removed the Snubber capacitance from both the TRIAC and the optocoupler.  Now the LED is perfectly off for nearly 1 hour.

Should I do the second part with opto completely removed.

[getfast_kiran]

When I removed the snubber capacitance of the optocoupler. I also shorted the 2.4k resistor part. Hence only that  resistance R is present in the optocoupler triac path
to the gate of the Driven Triac.

[getfast_kiran]

As a test, can you try the circuit without the snubber cap?
Without the opto & add a resistive load on the gate to ensure it's off...

As you told I removed the Snubber capacitance from both the TRIAC and the optocoupler.  Now the LED is perfectly off for nearly 1 hour.

Should I do the second part with opto completely removed.

Also since I plan to switch inductive loads also with the same...I will need this snubber circuit to the optocoupler. So is there a solution that might fulfill both the inductive load problems and LED bulb problems.

[getfast_kiran]

https://www.mathiselectronics.com/amp-clamp

How is the Amp-Clamp better? The Amp-Clamp shunts all leakage current away from the load, thus completely resolving all the above listed problems.  It is also multi-volt.  It can be used with controllers of any voltage from 24 to 120v.  Simply insert the Amp-inator into your output circuit and the leakage current no longer appears at the load.  Also, the Amp-Clamp only increases the “on” mode energy consumption by 0.14W, and it doesn’t get hot.  Seriously!  It operates at only a few degrees above ambient temperature.

found this as a solution of 120V AC only.

[getfast_kiran]

re you talking about a  you build it LED or a commercial LED LAMP.  I use <19W E27 lamps as night lights sometimes by putting a .1uf or smaller in series with the lamp.  Depending on the electronics inside, I get a nice low power lamp, it flashes, or doesn't turn on at all. All in series triac circuits have leakage issues. A good size RC (100ohm .47uF) across the lamp may solve some issues.

I will retry with the RC of 100 ohm and 0.47uf. But the thing is as fellow eevblog forum mate told me When I removed the snubber all together the problem got solved and there is not flickering or flashing.
But I need this circuit to operate with inductive load. So I will need a snubber circuit eventually.  So do you think this particular value of snubber will help me with the issue.

[capt bullshot]

Is there a reason why you need to use the triac? A simple relay would do the job fine, I believe.

[madires]

You can get snubberless TRIACs and snubberless TRIAC driver optocouplers:
- http://www.st.com/content/ccc/resource/technical/document/application_note/ca/ef/9d/1e/4a/9f/47/69/CD00003865.pdf/files/CD00003865.pdf/jcr:content/translations/en.CD00003865.pdf
- https://www.fairchildsemi.com/collateral/Snubberless-TRIAC-Drive-Optocouplers.pdf

[Seekonk]

I will retry with the RC of 100 ohm and 0.47uf. But the thing is as fellow eevblog forum mate told me When I removed the snubber all together the problem got solved and there is not flickering or flashing.
But I need this circuit to operate with inductive load. So I will need a snubber circuit eventually.  So do you think this particular value of snubber will help me with the issue.

The snubber should be across the LOAD.  That absorbs the leakage from the triac circuit.

[BrianHG]

You can get snubberless TRIACs and snubberless TRIAC driver optocouplers:
- http://www.st.com/content/ccc/resource/technical/document/application_note/ca/ef/9d/1e/4a/9f/47/69/CD00003865.pdf/files/CD00003865.pdf/jcr:content/translations/en.CD00003865.pdf
- https://www.fairchildsemi.com/collateral/Snubberless-TRIAC-Drive-Optocouplers.pdf

These solutions wont work since they still have a resistor/cap snubber across the output triac, even if they don't have one across the opto-triac output driving the power triac's gate.  It is the current leaking across this cap and resistor which eventually builds up a charge in the LEDs power supply making it pulse on, then discharge it's internal cap.

There isn't much that can be done other than getting rid of the common snubber all together and finding an alternative route to switching the AC.  Now, I have a method using a power mosfet and AC-DC bridge rectifier, but, it's something like a 10 component solution and you need to switch the optocoupler on only during the 0 crossing phase of the AC & you also need both access to the hot and neutral.

[Zero999]

As a test, can you try the circuit without the snubber cap?
Without the opto & add a resistive load on the gate to ensure it's off...

As you told I removed the Snubber capacitance from both the TRIAC and the optocoupler.  Now the LED is perfectly off for nearly 1 hour.

Should I do the second part with opto completely removed.

Also since I plan to switch inductive loads also with the same...I will need this snubber circuit to the optocoupler. So is there a solution that might fulfill both the inductive load problems and LED bulb problems.

If you can't put the snubber across the load, then you need to put a transient over voltage protection diode in parallel with the TRIAC. Hopefully it won't leak enough current to cause the LED to light.

[madires]

These solutions wont work since they still have a resistor/cap snubber across the output triac, even if they don't have one across the opto-triac output driving the power triac's gate.  It is the current leaking across this cap and resistor which eventually builds up a charge in the LEDs power supply making it pulse on, then discharge it's internal cap.

STMicroelectronics Application Note 437: http://www.victronics.cl/index.php?option=com_docman&task=doc_view&gid=110&Itemid=4

[getfast_kiran]

Yes I wanted to use Triac expecting it to deliver more life than the HIGH current mechanical relays. Do you think it justifies the cause.

[capt bullshot]

Yes I wanted to use Triac expecting it to deliver more life than the HIGH current mechanical relays. Do you think it justifies the cause.

Can't tell that without knowing your constraints. In general - it depends.
Mechanical relays are quite often used for such purposes, if chosen correctly I'd expect a life of >= 20 years and >= 100000 switching cycles for a standard power relay.
For example - the relays controlling the gas furnace and water pumps in my home heating system are switching on and off  10 ... 100 times per day and is expected to do this for more than 20 years.
If handled incorrectly, the relay may fail at first turn-on. A too large inrush current can weld the relay contacts together, exceeding the max. specified load may burn the contact at turn-off.

[Zero999]

Yes I wanted to use Triac expecting it to deliver more life than the HIGH current mechanical relays. Do you think it justifies the cause.

As stated above, a TRIAC isn't guaranteed to last longer than a mechanical relay.

Mechanical relays are expected to last a certain number of cycles before the contacts wear out. Semiconductor switches, such as TRIACs don't have that limitation, which makes them attractive in applications with more frequent switching than relays. Semiconductors are also more resistant to mechanical shock and vibration than mechanical switches.

On the other hand, mechanical switches, such as relays, can be electrically more robust than semiconductors. They wont be destroyed by ESD and can withstand over voltage and over current for longer than semiconductors. It's also possible to build relays which work at higher temperatures and are more resistant to ionising radiation than semiconductors.

[getfast_kiran]

Thank you both for that huge helpful info on mechanical relays. I forgot to mention the requirement of low height clearance of the circuit and space requirements.

So could you guys please help with the same without the need to replace the Triac ??

[getfast_kiran]

You can get snubberless TRIACs and snubberless TRIAC driver optocouplers:
- http://www.st.com/content/ccc/resource/technical/document/application_note/ca/ef/9d/1e/4a/9f/47/69/CD00003865.pdf/files/CD00003865.pdf/jcr:content/translations/en.CD00003865.pdf
- https://www.fairchildsemi.com/collateral/Snubberless-TRIAC-Drive-Optocouplers.pdf

I am using a snubberless TRIAC but the optocoupler is not snubberless will try it with a snubberless optocoupler in the future. For the time being
please do help me with solution confined to snubberless TRIAC BTA16 and non snubberless optocoupler 3021.

[getfast_kiran]

I will retry with the RC of 100 ohm and 0.47uf. But the thing is as fellow eevblog forum mate told me When I removed the snubber all together the problem got solved and there is not flickering or flashing.
But I need this circuit to operate with inductive load. So I will need a snubber circuit eventually.  So do you think this particular value of snubber will help me with the issue.

The snubber should be across the LOAD.  That absorbs the leakage from the triac circuit.

Could you give or point me towards a schematic that contains a snubber across the LOAD. Is the value of the snubber circuit the same.

[getfast_kiran]

You can get snubberless TRIACs and snubberless TRIAC driver optocouplers:
- http://www.st.com/content/ccc/resource/technical/document/application_note/ca/ef/9d/1e/4a/9f/47/69/CD00003865.pdf/files/CD00003865.pdf/jcr:content/translations/en.CD00003865.pdf
- https://www.fairchildsemi.com/collateral/Snubberless-TRIAC-Drive-Optocouplers.pdf

These solutions wont work since they still have a resistor/cap snubber across the output triac, even if they don't have one across the opto-triac output driving the power triac's gate.  It is the current leaking across this cap and resistor which eventually builds up a charge in the LEDs power supply making it pulse on, then discharge it's internal cap.

There isn't much that can be done other than getting rid of the common snubber all together and finding an alternative route to switching the AC.  Now, I have a method using a power mosfet and AC-DC bridge rectifier, but, it's something like a 10 component solution and you need to switch the optocoupler on only during the 0 crossing phase of the AC & you also need both access to the hot and neutral.

Do you have a schematic of the 10 component solution. I would like to test it if there is no other solution since I will have to light a LED lamp eventually.

[getfast_kiran]

As a test, can you try the circuit without the snubber cap?
Without the opto & add a resistive load on the gate to ensure it's off...

As you told I removed the Snubber capacitance from both the TRIAC and the optocoupler.  Now the LED is perfectly off for nearly 1 hour.

Should I do the second part with opto completely removed.

Also since I plan to switch inductive loads also with the same...I will need this snubber circuit to the optocoupler. So is there a solution that might fulfill both the inductive load problems and LED bulb problems.

If you can't put the snubber across the load, then you need to put a transient over voltage protection diode in parallel with the TRIAC. Hopefully it won't leak enough current to cause the LED to light.

Could you please send me the specs of the Transient Over voltage protection diode. ... to be added in parallel to TRIAC.

[getfast_kiran]

These solutions wont work since they still have a resistor/cap snubber across the output triac, even if they don't have one across the opto-triac output driving the power triac's gate.  It is the current leaking across this cap and resistor which eventually builds up a charge in the LEDs power supply making it pulse on, then discharge it's internal cap.

STMicroelectronics Application Note 437: http://www.victronics.cl/index.php?option=com_docman&task=doc_view&gid=110&Itemid=4

Why did you give reference to the particular application note.... I did not understand the reason behind it. Could you please explain.

[capt bullshot]

For a simple attempt:
leave your circuit as it is and place a resistor from its output to the neutral line. This should divert the snubbers leakage from the LED lamp.

In other words: Put that resistor (for example 22k, 3W, 250V rated) across the lamp. That should stop the flashing for the price of significant power dissipation in that resistor (you do not want that, but it is a quick fix).

For a better attempt:
change the values of the 100n caps (C9, C4) in your circuit to 10n. These cause the leakage that causes the LED lamp to flash. If the LED is still flashing, put a 100k Resistor across the lamp. This way you'll have less power dissipation and smaller components, but you need to check if the snubber is still effective enough.