Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Reason of frequently triac failure in the dimmer

Status
Not open for further replies.

Willen

Well-Known Member
Hi,
I prepared a dimmer for colorful lights which includes some transformerss colorful LED lamps and some (approx 100) colorful tungstan small bulbs (total wattage is around 80W). Thus I think the entire load is not 'inductive'. So I didn't add RC snubber to protect Triac in the dimmer. It would be compulsory if I used inductive loads instead. Look at the attached circuit.

Problem:
1) First I used 2N6073A (sensitive gate 4A, 400V) triac and it worked one night, but failed to another night. Failure: I am not being able to dim the lights, their brightness are full always. The damaged triac has unusual 'G to A1 short'.
2) Then I replaced with BT136 triac and it worked one night, but failed to another night again. Failure: I am not being able to dim the lights, their brightness are full always. The damaged triac has unusual 'A2 to A1 voltage drop of around 600mV' as diode.

What would be the reason of the frustrating problem? I used NTC, mains induction filter (T1) and X capacitor just for fun there. :) Do you think the T1 causing some inductive effect to destroy the triac again and again?
 

Attachments

  • DIMMER.GIF
    DIMMER.GIF
    5.7 KB · Views: 577
I am not good with triacs. Try R2, C1 and see what happens. (across triac) That should help with inductive kick.
I don't think your filter adds "kick" because Cx should keep that side from flying up.
**broken link removed**
 
Hi Willen,

You need a 470 Ohm resistor between the DIAC and the TRIAC gate to limit he TRIAC gate current when the TRIAC fires. This will also protect the DIAC from conducting excessive current

You should also connect a 4.7K Ohm resistor between the TRIAC gate and neutral. Otherwise the gate is floating when the TIAC is not conducting and this can lead to odd TRIAC behaviour.

The TRIAC needs to be mounted on a heatsink. An A1 to A2 short is a typical symptom of fusing caused by overheating. The TRIAC and heatsink need to be in free air. (the maximum junction temperature of the 2N6071 is 110 Deg C and the BT1216 maximum junction temperature is 120 Deg C)

You say that the load is 80W and that includes tungsten lamps. The cold resistance of tungsten lamps can be only one tenth of their hot resistance so you need to take this into account when calculating the TRIAC peak current. Also are you sure that the load is 80W. Connect the load to the mains without the dimming circuit and measure the actual current that the load takes. Observe safety precautions.

spec

PS: You may think that this circuit is fine because it is on the net. Many circuits on the net are badly designed, or worse still, could never work. The other thing is that sensitive gate TRIACS can be a bit fragile.:)

DATASHEETS
https://www.onsemi.com/pub_link/Collateral/2N6071-D.PDF
https://www.farnell.com/datasheets/1758172.pdf?_ga=1.201288102.942089101.1451155200
 
Last edited:
If you have transformers in the circuit then its definitely inductive, rons suggestion is a good one.
Something else important if your using transformers is to have a circuit that triggers symmetrically, the circuit above isnt much good for transformers in that respect.
 
T1 in Willens circuit is a losey anti phase EMI transformer, intended to absorb energy. Although it looks lie an ordinary transformer on the schematic its inductive effect is negligible.

spec
 
OK, yes maybe it can work without failure as suggested, with RC snubber, Gate to neutral and diac to Gate resistors.

Is there any clue why a tiny TO-92 triac can control hundreds of watts easily? It's awesome!
 
Is there any clue why a tiny TO-92 triac can control hundreds of watts easily? It's awesome!

Mmm- neither of the TRIACs you mentioned in the OP are in TO92 cases. Perhaps you are thinking about another TRIAC.

The reason why switching components: diodes, SCRs, TRIACS, BJTs, IGJTs, MOSFETs, can handle so much power is because they are either off or on. Power in Watts is V* A. So if the component is switched off there is no current so V * A = V* 0 =0 Watts. When the component is turned on there is dissipation. Take a diode: it typically has a forward drop of 1V so the power dissipation, when it is conducting, is 1V * A. If A is 0.5A the power dissipation would be 1V * 0.5A = 0.5W. If the voltage being switched were 1000V direct, the power being switched would be 1000V * 0.5A = 500W.

A TRIAC forward Voltage is typically 2V so a TO92 TRIAC would be able to conduct 0.5W/2 Amps = 0.25A. So if a TRIAC were switching a 1000V voltage the TRIAC would be controlling 0.25A * 1000V = 250W

On the other hand a TO220 TRIAC, mounted on a big heatsink, would be able to dissipate 30W maximum. So the maximum current would be 30W/2V = 15A. Once again, if you had a 1000V voltage supply you would be able to switch 15A * 1000V = 15000W.:cool:

spec
 
Last edited:
I Believe your 5.6K resistor is TOO LOW in Value, Supplying Too Much Gate Current.
Try the Circuit in Post 2.
 
Thank you for the more clarification. I attached an equivalent transistor circuit of triac. Do it get fully workable triac with two PNP and NPN general purpose transistor as shown in attachment? Maybe not due to low current and low voltage rating of the transistors. But probably I would get low voltage triac (SCR) like around 40V in this way with general purpose transistor.
 

Attachments

  • 220px-Triac_Quad_I_like_SCR.JPG
    220px-Triac_Quad_I_like_SCR.JPG
    8.2 KB · Views: 314
That may be Somewhat Equivalent, But I wouldn't try it, especially Not on Line Voltages.
And the Two Transistors Must be Complementary Pairs.
 
Hi,
I prepared a dimmer for colorful lights which includes some transformerss colorful LED lamps and some (approx 100) colorful tungstan small bulbs (total wattage is around 80W). Thus I think the entire load is not 'inductive'. So I didn't add RC snubber to protect Triac in the dimmer. It would be compulsory if I used inductive loads instead. Look at the attached circuit.

Problem:
1) First I used 2N6073A (sensitive gate 4A, 400V) triac and it worked one night, but failed to another night. Failure: I am not being able to dim the lights, their brightness are full always. The damaged triac has unusual 'G to A1 short'.
2) Then I replaced with BT136 triac and it worked one night, but failed to another night again. Failure: I am not being able to dim the lights, their brightness are full always. The damaged triac has unusual 'A2 to A1 voltage drop of around 600mV' as diode.

What would be the reason of the frustrating problem? I used NTC, mains induction filter (T1) and X capacitor just for fun there. :) Do you think the T1 causing some inductive effect to destroy the triac again and again?

Hi,

A 400v rating on a triac used for 220vac is not really good enough either. That would be better on a 120vac system I always used the highest i could get even on a 120vac line like 600v.

Try a 600v or higher rated triac instead. There are sometimes surges and spikes on the line that we dont often think about. I worked briefly at a company that made high quality power line monitors and the tests for those things you would not believe.

What is that "NTC" for?

Compare your circuit to other circuits like that one on the web.
 
That may be Somewhat Equivalent, But I wouldn't try it, especially Not on Line Voltages.
And the Two Transistors Must be Complementary Pairs.

Try a 600v or higher rated triac instead. There are sometimes surges and spikes on the line that we dont often think about.
What is that "NTC" for?
I saw 'NTC' in a thermistor (used in series of mains of a computer SMPS)

100 meter far, my neighbour have a big 3 phase motor. When they turn ON and OFF, I can feel the effect in my home in mains. Do they generate spikes and can affect to my triac of my home?
 
I Believe your 5.6K resistor is TOO LOW in Value, Supplying Too Much Gate Current.
Try the Circuit in Post 2.
The circuit of post #2 will also overcurrent the DIAC and TRIAC. The relatively high value phase angle capacitor is like a zero ESR battery. The current will be a short pulse but still enough to do damage. The problem may not be apparent for a while but as the overcurrent occurs 50 times a second sooner or later there will be a failure, especially as the DIAC and TRIAC junction temperatures increase.

To sort this circuit you need to increase the value of the 5.6K resistor, but you also need two additional resistors, as stated in post #3.

spec
 
Last edited:
Thank you for the more clarification. I attached an equivalent transistor circuit of triac. Do it get fully workable triac with two PNP and NPN general purpose transistor as shown in attachment? Maybe not due to low current and low voltage rating of the transistors. But probably I would get low voltage triac (SCR) like around 40V in this way with general purpose transistor.
You can make a TRIAC from an NPN and PNP transistor. In fact, it is an SR flipflop if you put a resistor in series and that configuration is used widely in circuits, especially integrated circuits.

But as it stands, the maximum current is limited to the maximum base current of the transistors- not very much. Although, you can include a few resistors so that the maximum current limit is the maximum collector current of the transistors.

There is a major problem though, and that is the fusing current which is dependent on temperature and rate of rise of the current di/dt. SCRs, TRIACs etc are specially designed for high fusing currents.

And, if all that is not enough, there is also the problem of false triggering.:D

spec
 
Last edited:
Hi Willen,
The NTC resistor does not protect against external spikes. It limits the current taken by the device at switch on. This occures with switch mode power supplies as the main input capacitor has to be charged at switch on. It can also occure with transformers (Particularly toroidal transformers.) if the instsntanious voltsge at switch on is trying to magnetize the core in the same direction it was left magnetized in when it was switched off. The core will be near saturation so there can be no flux change to generate a back EMF. I had to fit one to a 200 VA lighting transformer to stop it from tripping the 6 amp MCB on the lighting circuit. The NTC resistor has a high resistance at switch on and is heated up as a result of the current passing through it. After a second or two it's resistance drops so it does not drop the voltage to the equipment very much. Adding a varistor after the NTC or an ordinary resistor would provide better protection from external voltage spikes.

Les.
 
The antiphase lossy transformer/inductor on the input also protects against spikes on the mains.

spec
 
Hu Willen, I remember you saying your supply is very limited, its quite possible a 3 phase motor on the same grid is generating noise problems.
Maybe try a transorb to clean up the mains.
 
So why not just buy a heavy duty rated dimmer and be done with it? Why reinvent the wheel and do it poorly? o_O
 
Theres no factory made light dimmers for sale in Nepal? o_O
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top