Driving two triacs with one triac driver

Hi,

I had made a circuit in which a set of 4 halogen lamps were driven by a triac ( to control dimming)

However it resulted in frequent triac burn outs ( even though the current takes only 15A and the triac is rated for 40 ), so what I did is, I have now redone the circuit with two triacs, each driving 2 lamps. This would not result in any burn out.

I have attached my schematic below. Please let me know if I have done it right. Thanks!

This can't work reliably. Triacs with separate MT1 connection need separate gate drivers.

Hi,

A triac (bjt, mosfet or any other part) burns because of
* over current
* over voltage
* over heat (caused by power dissipation)

Your traic has no overvoltage protection, thus every series inductance (filter inductance, transformer, wire inductance) may caus high voltage spikes. They may be short in time, but they will hurt a semuconductor. Not necessarily immediately, it may be after a couple of seconds or a couple of months.

Overcurrent: we don't know which load there is connected. Maybe a transformer. If so, then every tiny mismatch in control angle of the positive to the negative halfwave will cause high DC currents. Too high for a triac. Sometimes you can hear the transformer humming when this happens.

Overtemperature: we don't know about a heatsink. You need to take care of temperature. With two triacs you won't significantly reduce overall power dissipation, but now it's divided into two triacs. This is an improvement, but you still may need a suitable heatsink.

Schematic: it should work, but I'd rather used two resistors - one at each gate.

It seems you don't like capacitors. There is not a single one. Not as bulk capacitor, not for decoupling, not for filtering not for a snubber.

And we don't see your PCB layout. Dangerous high voltage combined with low control voltage is always critical. There are safety regulations. Especially take care about creepage and clearance distances.

Klaus

Hi,

thanks for the reply. Would it be ok to do it with 2 MOC3021s ?

The load connected are halogen lamps ( 1KW each )

What could I do about overvoltage protection and adding capacitors. Need a bit of help here, I'm not a professional electrical engineer, just a hobbyist with some experience.

Hi,

two MOCs:
As said: It should work the way you do. Don´t focus on that.
But to answer your question:
It will be OK with
* your circuit,
* with two independent Rs and one MOC and
* with two MOCs.

Halogen lamps:
Read about halogen lamps. Especially the high inrush current when the filament is cold.

Triac overvoltage protection:
There are more than enough discussions here in the forum and in the internet. Every triac manufacturer will provide application notes about triac protection.
Read through them, then try to find a solution for you and - if you´re in doubt - show us.

I´d say more than 99% of the circuits need capacitors. Thus you should be warned if you see a circuit without a single capacitor.
Arduino surely needs a stable power supply. Stable means, thet there should not me much ripple. Low frequency ripple is canceled out by a single bulk capacitor, and high frequency ripple is cancelled out by ceramics capacitor at every load that draws pulsed currents.
Whether you need some or not is written in the datasheets. The point is: you - as the designer - are responsible. Even if it is not explicitely written that you need capacitors, you have you are responsible to satisfy the specificatons in the datasheet. Some maybe say: 5V +/- 5%. Then this demand is not satisfied when your DC voltmeter shows 4,9V.
Because your DC voltmeter shows an average value. But you need to guarantee that the voltage is not outside the specified range for a tiny piece of a second. Not even for 10ns.
Simply the traces from your AC/DC supply (btw: we know nothing about it) to the arduino may cause enough inductive impedance that there are tiny negative peaks --- outside specified values. A 100nF capacitor close to the arduino may be sufficient.
For sure the arduino itself may have enough capacitors - I don´t know.
But then there is a fan .. additionally causing current spikes... (again here we know nothing)

But usually we dont need to know this all. You need to read the datsheets.
I know I repeat myself to read datasheets. This is because it is so urgent. We all need to do this. The professionals need to read even more datasheets, regulations...
I know this is boring, but there is no way around.


Klaus

Hi KlausST. I am sorry for not supplying complete information and apologies if I come off as too demanding. I have come to realize that electrical/electronics design is a deep field and I need to do my research properly to come off with really good circuits. Hopefully this forum will help me to improve.

The AC-DC converter is a hilink solid state device 0.41 A 5W
( https://www.sunrom.com/p/ac-dc-isolated-power-module-220v-to-12v041a-5w-hlk-5m12 )
The fan is a normal brushless 12V DC fan used to keep the electronics cool.

I read about RC snubbing. But I can't seem to get the values right ( ie. what should be R and C ). The calculations seem to depend on dV/dt, parasitic inductance and such like ( which I don't have any idea what they should be ).

I have redesigned the circuit to use just the 1 triac but with capacitors for snubbing. Does it look better?

Hi,

Now you changed back to one triac.
I don't know if this is correct or not.
* Did you read about the halogen lamp inrush current? What peak current can you expect?
* Did you verify it with the triac specifications?
* Did you choose proper heatsink for the expectable triac power dissipation?

Klaus

Hi,

The halogen lamps are 3.3A. Four of them means about 15A maximum. The triac is rated for 40A. So it really should work with just one. And in fact, the board I made before does work with just one, it just kept burning out now and then. That's why my 'solution' was to have 2 triacs, but it occurred to me that the solution could be something else.

About the inrush current. I read that it can be about 10x the normal running current. Meaning about 150A. Now I suppose the capacitor thing is there to limit the current in some way ( but I m not sure how it works ).

What do you mean by heatsink? I have a small aluminium one that attaches to the back of the triac, and a fan to keep it cool. It doesn't get heated up much. In any case the burnouts don't happen when it is being used for long time, they happen when it just gets powered on. So I suspect the inrush current is to blame.

Hi,

The halogen lamps are 3.3A

Click to expand...

In my eyes impossible.
You say 230V and you say 1kW each...
But 1000W/230V = 4.34A RMS...
Then there is the problem with the cold filament
And the problem that the peak is 1.414 times the RMS value

I won't be surprised if the cold peak reaches 30A.

Added:

Now I suppose the capacitor thing is there to limit the current in some way ( but I m not sure how it works ).

Click to expand...

I wonder how you come to this conclusion.
In short: It is completely wrong.
A snubber usually is for keeping dV/dt low... and it may keep any inductive feedback voltage low.
Mind: both are about voltage ... they have about nothing to do with current.

What do you mean by heatsink?

Click to expand...

The triac will dissipate electrical power/energy into heat.
Even with your given 15A it will be too much heat to be spreaded with a small piece of aluminum.

Klaus

I agree with the calculation of current. The 3.3 A was what was written on the tube, but then it also said 1kW. Anyway, let's say that the peak current is 25A. So, how do we control the current inrushing?

Specifications of BTA41 are as follows

I_rms (on state rms current) : 40 A
I_tsm (Non repetitive surge peak on-state current ) : 400 A
I²t Value for fusing: 1000 A^2 sec
etc.

Unfortunately, the datasheet doesn't include an example circuit.

Hi,

I read that it can be about 10x the normal running current. Meaning about 150A.

Click to expand...

let's say that the peak current is 25A.

Click to expand...

25A per tube...
We can not verify the tube specifications...we need to rely on your values.

Then the triac should work (I´m not completely convinced.) Additinally you have 4 outputs, but you speak about 2 tubes.

Thermal considerations.
According Figure1 you may expect about 5W of power dissipation at 6.6A RMS 180° degree.
According Rth(j-a) you may expect a 50K/W x 5W = 250K of temperature rise.
With 40°C ambient temperature this is 290°C.

To ensure a junction temperatur of less than 125°C you need an Rth(j-h) of (125°C -40°C) / 5W = 17K/W.
Subtract the 0.9K/W Rth(j-c) and get a Rth(heatsink) of about 16K/W.

Look at heatsink manufacturers for heatsink with comparable Rth just to get a clue about size.

Klaus

There are 4 tubes, with 4.34 A per tube rms (as per prev. calculation). That comes out to 17A total RMS and 25A peak. That is, in the running condition. During the inrush, current can be 10 times of this. We must try to limit this current, right?

There seem to be some thermistors available which can limit the current., eg. Ametherm SG379. Would it be a good idea to connect one of them to the input power supply?

I will think about the thermal considerations.

Hi,

sorry I frogot to answer this:

So, how do we control the current inrushing?

Click to expand...

If the triacs are able to carry the load: Then it should be sufficient to slowly ramp up the phase angle.
Then with moderate current the filament is able to heat up.

BTW: some professional lighning control don´t switch off the lights completely. They keep the filament hot, but very dim.
This keeps next inrush current low and faster to get bright light.

Klaus

I will take your advice to slowly ramp up the phase angle and see if it gets burnt out again. If it does, I will add the thermistor current limiter just to make sure.

Not switching off the light is not an option in this case, in fact they will be off most of the time, and will need to be switched on.

Hi,

Can I ask if using a DC plug 2.1 mm dia barrel plug for AC power supply ( 230 V 4 A ) is a good idea? Otherwise, what kind of connector would be preferred?

Hi,

What does the datasheet say?

Klaus

Unfortunately, there isn't any datasheet as such. It's just a common DC barrel plug that I got from here
**broken link removed**

I'm just worried that the high voltage AC may cause the plug to break down and get short circuited or something.

saurabhsaurc said:

Can I ask if using a DC plug 2.1 mm dia barrel plug for AC power supply ( 230 V 4 A ) is a good idea?

Click to expand...

This type of connector usually has small solder tabs and a small compartment to fit the wires into. 4 Amps is a lot to send through the wires which typically go to the terminals of this type of connector. You can solder them securely, and the plug can maintain its insulation qualities, however over time the wires are likely to break. Then you may have 1kW of power short-circuiting.

Ok thanks, I'll go for a different connector!

Hi,

230V is dangerous, thus you have to keep on safety regulations.
Don't use a single device on dangerous voltage where you don't have a datasheet.

Klaus