02 Laptop adapters kill BTA16-600BRG triac on turning off

[paulsteigel]

Hi Everyone,
I found a thread asking on inductive load for Triac can kill the triac with clamping voltage, however, the thread is no longer open so let's me make a question here and hope you will be so kind to help me!
I made a relay using a BTA16-600BRG with MOC3020. The relay working find with normal lamp, and less than 500W devices including fans (with capacitor), drilling tool. As datasheet tells, BTA16 is a snnubless triac so I decided not to make R-C snubbed circuit.
I made a final test with switching 02 laptop adapters (about 400W totally). The triac died right at my first switching off. A1 and A2 remained totally open with almost 0ohm measuring between A1 and A2.
I knew cheap Chinese made Triac may not be as good as the genuine one so I doubt that my BTA16-600BRG does not have built-in snubbed circuitry and the two laptop adapters have killed the triac.
So I would like to have your advice:
1.What caused the triac death?
2. Can I build a R-C snubber circuit to protect the triac?

I also attached a new version of my triac with snubber circuit and preparing to do some testing for now. The original one does not have this protection.
Thank you very much for your supports and advice.
Old schematic:

and here is my new version and i am starting to build a typo for testing.

 

[betwixt]

What are U5 and U6 for?
You have the connections to the MOC3020 wrong on the new schematic and on both schematic there is no supply to the optocoupler LED.
Consider that the MOC3020 is a random phase device, it is possible to turn the triac on at peak AC voltage which might cause a high current to flow, for AC load switching you might find the MOC3040 more appropriate as it works identically but has built in zero crossing detection.

Brian.

 

[FvM]

How do you recognize that the triac "died" at turn-off and not already failed short circuit on initial inrush current?

Inductive overvoltage can cause overhead self-triggering but doesn't easily damage a triac if the inductive energy is limited. Exceeding the I^2t rating will fastly do.

A snubber is always recommended with inductive load to avoid unwanted dv/dt triggering. But it doesn't help against overcurrent induced damage. A zero crossing detector triac driver will hopefully prevent from switching into high inrush currents of a load like SMPS input capacitors.

 

[paulsteigel]

What are U5 and U6 for?
You have the connections to the MOC3020 wrong on the new schematic and on both schematic there is no supply to the optocoupler LED.
Consider that the MOC3020 is a random phase device, it is possible to turn the triac on at peak AC voltage which might cause a high current to flow, for AC load switching you might find the MOC3040 more appropriate as it works identically but has built in zero crossing detection.

Brian.

Sorry for replying late, U5,U5 is the two SN74LVC1G3157DBVR to act like a swaping switch so that turning off the relay by Touch module or by MCU can toggle the relay. This is a way of applying stair case switch to the relay.
My second version has a schematic problem (Connecting to Pin4 instead of 5-NC). My actual test was correctly done with 4 and 6 on MOC3020.
Supply to MOC3020 is 3.3V DC from:
1. Touch module TTTP223 >> about 3.1V and
2. From ESP8266 TX pin >> abour 3.3V
This work fine on actually test, the only things I have problem is when connecting 02 laptop adapters (about 480W totally), the Triac died on switching off, A1 and A2 connected immediately and does not release.
Finally, I found a the problem is from the Triac itself. I bought 2 genuine BTA16-600B at double price to the one I used (bought from China and died) and it worked fine with 2 adapter and a 440W pump (without snubbed circuit).
Today I decided to make a test build of subber circuit for the Cheap CHinese BTA16-600B and it work fine with 2W-390Ohm resistor, 22nf/600V capacitor connected in series.
As my purpose for the relay is just to turn on/off lamp: normal AC 220V/upto 240W bulb / Led, Tungsten so test on serious inductive load is by far to see how far the relay can go.
The solution is: making clear warning so user will not load the relay with adapters or device with high inductance/ capacitance.
Thank you very much for your attention!

--- Updated Jan 15, 2021 ---

How do you recognize that the triac "died" at turn-off and not already failed short circuit on initial inrush current?

Inductive overvoltage can cause overhead self-triggering but doesn't easily damage a triac if the inductive energy is limited. Exceeding the I^2t rating will fastly do.

A snubber is always recommended with inductive load to avoid unwanted dv/dt triggering. But it doesn't help against overcurrent induced damage. A zero crossing detector triac driver will hopefully prevent from switching into high inrush currents of a load like SMPS input capacitors.

The triac did not act on controlling, just having A1 and A2 connected fulltime Sir!

 

[FvM]

The triac did not act on controlling, just having A1 and A2 connected fulltime Sir!

I know. I presume it failed short when turned on for the first time with the power supply load. But you don't become aware of before trying to turn it off again.

 

[paulsteigel]

I know. I presume it failed short when turned on for the first time with the power supply load. But you don't become aware of before trying to turn it off again.

Thank you, I realised the triac death at pressing off button as low wattage resitor start to burn but at turning on/ nothing happened (when install a R-C snubber circuit). That would means with initial test, no snubber circuit, the triac died on switching off High Capacitive load (02 adapters case).
I make to separated tests for getting known to safety margin of the product.
Condition applied: no heatsink, no snubber circuit
Test 1: Resitive >> all fine, up to 2A for bulbs (Incandescent light bulb, Fluorescent lamp and Led driver) up to 100W;
Test2: Inductive >> All fine two with devices including 02 normal fans, 01 drilling tool - 710W
almost 900W
Test 3: 01 bulb and 01 Laptop Adapter (DELL E5440) >> Ok
Test 4: 01 bulb and 02 laptop adapters (HP and Dell) >> failed
So with my original expectation of switching on/off light bulb, electromagnetic contactor for switching AC high power device, I think even the cheap Chinese BTA16-600B has successfully accomplished its job. Except for failure of powering high capacitive load like laptop adapters (that the genuine BTA16-600BRG from ST has overcome successfully).
Measurement: Clearly note users on proper use of the relay >> not to use with powering high power device.

PS: I tested with a subbed circuit(C:22nf,630V, R:330,2W - triac side; R150,2W - Optocoupler side) and the Cheap Chinese Triac worked jusst fine with 02 Laptop adapters except little leakage that makes bulb dimming a little bit.

 

[KlausST]

Hi,

Best is to measure actual conditions. Like load current and load voltage.
Then you will see what really distroys the triac.

A light bulb will draw maybe 10 times inrush current (when cold). Switching capacitors directly with a triac may be no good idea at all.

Generally semiconductors get killed by:
* overcurrent (peaks) (usually burnt bonding wires --> fail to become "open", )
* overvoltage, even very short spikes (distroyed isolation barrier causes increased isulation current which causes a lot of heating power . Short time or long time failure --> often fail to become short circuit)
* overtemperature of the die (ambient temperature + self heating --> often fail to become short circuit)

Overtemperature of die may also happen within milliseconds. See SOA chart.

Klaus

 

[paulsteigel]

Hi,

Best is to measure actual conditions. Like load current and load voltage.
Then you will see what really distroys the triac.

A light bulb will draw maybe 10 times inrush current (when cold). Switching capacitors directly with a triac may be no good idea at all.

Generally semiconductors get killed by:
* overcurrent (peaks) (usually burnt bonding wires --> fail to become "open", )
* overvoltage, even very short spikes (distroyed isolation barrier causes increased isulation current which causes a lot of heating power . Short time or long time failure --> often fail to become short circuit)
* overtemperature of the die (ambient temperature + self heating --> often fail to become short circuit)

Overtemperature of die may also happen within milliseconds. See SOA chart.

Klaus

Tks Klaus, I am doing more tests for stability and will try to take your advice into a next version with thanks!