BTA12 Triac Dimmer for Bulbs

[manoj.patil.1974]

Hi, I have spent sleepless nights on this simple circuit that refuses to work. Can some one please help with with what's going wrong here ?

I am using a 230V A/C, 100W filament bulb as a load.

-> S2 on , bulb is not switched.
-> S2, S3 on, bulb is on and dimmed
-> S2, S3, S4 on, bulb is on and more brighter.
-> S1 on or off has no effect on the bulb brightness.

I googled and this circuit appears in many places and instead of R4,R5,R6 a 220K or 470K pot is used. However, in my experience, the bulb does not light up if R4+R5+R6 is more than 33K.
Also, R4,R5,R6 get hot. My question is why do I require so low value resitors (< 33K) for my circuit when everyone is able to handle it with 470K pot's ? I actually want to use a pot and want to have multiple dimmings settings.

R1 and R2 get very hot too. May be I should use 33K 1W or 15K, 4W resistors.




Here are my some debugging attempts
-> Instead of R4,R5,R6 I used a 16mm 470K pot and it smoked. I tried a 23mm 470K pot and it fumed too. Both were carbon pots. Should I try using 470K wire wound pots ? Problem is wire wound pots are not available for values greater than 100K
-> The capacitor has 100n written on it. I assumed its 100 nano farads. I even tried with a capacitor with 0.22 uf written on it. No change in behaviour
-> I have built this circuit atleast 5 times with FRESH NEW components.
-> Does it matter how I connect the Main Terminal 1 and 2 of the triac ?? Some websites state that it does matter but I thought these days triacs are reliably symetric.

Not sure what is going wrong here ? Are the BTA12 faulty or not behaving as per design ? These days the market is flooded with cheap chinese components hence its difficult to differentiate the real authentic ones with the fakes one's.

 

[Techie7]

A comic circuit. Better circuit are in internet. I think r3,r4,r5,r7 connected to downwards ie, the other supply terminal where bulb is not connected. 220K pot will go but a good Wattage is must. I can't understand need of D1,D2,D3,D4 and associated idiotic things wired.

 

[manoj.patil.1974]

Dear Techie7,

What is your exact point about R3, R4, R5, R7 ?

I would request you to avoid using the words "comic", "idiotic" because they are against the spirit of this forum.

The Diode network indeed has a purpose but now that you call it idiotic, I would avoid giving the reason. Anyways, there is a switch to put it out of the circuit.

 

[manoj.patil.1974]

More updates, more frustration.

I modified the above circuit for following combinations

1. Bulb was connected on the live wire of mains before connecting the triac circuit
2. Bulb was connected on the neurtal wire after the triac circuit
3. Bulb was connected before the MT1 of Traic
4. Bulb was connected after the MT2 of triac

No Luck

The results are really wierd (atleast for me)

E.g In the above circuit posted in my first post, I removed the bulb from MT2 of triac and connected it to MT1 ( i.e. on the lower side of the traic). The triac fires only with a 33K resistor. I then just swapped the wires connecting to Pin1, Pin2 of the triac and the bulb was glowing on with 50% brightness.

 

[em2006]

The sketch requires a small correction, see attach.

 

[manoj.patil.1974]

Dear em2006

Thanks a lot for posting the correction.

However, I have tried this combination too. With your suggested modification, result is
-> S2 on , bulb is not switched.
-> S2, S3 on, bulb is on and dimmed
-> S2, S3, S4 on, bulb is on and more brighter.
-> S1 on or off has no effect on the bulb brightness

I even tried to swap the input live and neutral wire, same result.

Then, with your modification wired, I just swapped the MT1 and MT2 of the triac. This time, the bulb on with approx 50% brightness. The bulb was on even when S1,S2,S3,S4 switches were off. Looks like the traic gets the gate current directly through capacitor. I opened the Gate wire and the bulb went off that proves that the Triac is not shorted.

There is something very fundamental that I am missing here about triacs

 

[em2006]

...-> S1 on or off has no effect on the bulb brightness...

S1 has an effect on hysteresis.
Dimmer curve hacking

 

[manoj.patil.1974]

Right. Its indeed for the hysterisis. But right now, I can't get the basics working so hysterisis is not yet connected.

 

[ccurtis]

I don't have the time right now to look at the core problem, but a quick calculation shows that R1 and R2 are dissipating 1.8 watts, higher than the 1 watt max they are rated for. That explains why they are hot. They would have to be 27K ohm to get them down to dissipating 1 watt, and resistors really should never be run at more than half their rating for good reliability.

 

[ccurtis]

What is the bulb brightness with S1 on and the others off?

 

[em2006]

Probably the original schematic was designed for line voltage 110v, not 230V.

 

[ccurtis]

Have you tried S2, S3, and S4 with the capacitor disconnected, and S1 off? If so, what is the result, for each switch combination (S2 thru S4)?

Roger, EM.

 

[manoj.patil.1974]

When S1is on and S2,S3,S4,S5 are off, the bulb does not light up at all.

I have not tried the last combination you have suggested with capacitor disconnected. I will try it and update you.

However, based on the feedback by all the readers here, I did following
1. Never switched S1 on and disconnected R1,R2 so as to isolate the hysterisis circuit
2. Connected the load on the Live wire, Neutral wire, MT1, MT2 in 4 different experiments, just to check if the positioning of the load was any way root cause.
The results are still wierd. For one positioning of MT1, MT2, the triac never fired with S1 on, S2 on made bulb light dim, S3 on made the bulb more brighter. When I reversed the Pin1 and Pin2 of triac, the bulb was always on irrespective of S2,S3,S4 positioning.
I observed this pattern for all the 4 different load positioning.

Removing capacitor is something I did not think.

Any idea what would be the resistor value that would fire a BTA12 600B near the peak of each half cycle ?


Note: I am using BTA12 600B and BTA16 600B for my experiments. These are NOT the snubberless version. Do you guys think not having a snubber in my circuit could be the problem. I don't think that is related as I have a simple resistive load, but I am a bit rustic on this now. Also, it would be great if you can point me to an arcticle to design/calculate snubber for my circuit just in case.

 

[ccurtis]

Any idea what would be the resistor value that would fire a BTA12 600B near the peak of each half cycle ?

Assuming your circuit is connected, properly after the lamp, and S1 is open, a total of 100K ohms with a 100nF capacitor will fire the diac at the peak of the 230 volt source. It does not matter what triac you use. The diac determines the turn on point. I determined this by simulation.

 

[ccurtis]

It is not usual in the dimmer circuits I have seen for R8 to be inserted. Is there any particular reason it is installed? Maybe that is affecting the ability of the diac to fire properly from the voltage on the capacitor.

If you remove the capacitor, just to try it, the lamp should be quite bright with any combination of the brightness switches.

 

[em2006]

When S1is on and S2,S3,S4,S5 are off, the bulb does not light up at all...

What about S5? Do you have added a new switch?
Why not use a potentiometer?

 

[manoj.patil.1974]

em2006,
Sorry about that S5. There is no S5 in my circuit

 

[manoj.patil.1974]

Ok.
Here are more results of my experiments.

1. I purchased a brand new BTA16 600B, 1W Resistors of value 33K, 100K, 200K
2. Simplified the circuit to just 1 Triac, 1 Resistor, 1 Bulb of rating 230V, 100W
3. Connected the Resistor to bulb and MT2. Other end of resistor to Gate. MT1 to neutral and other end of bulb to Mains live. For 33K, the bulb was on dim, for 100K and 200K, the bulb never turned on. For 15K, the bulb was more bright but not full on.
4. Modified the above circuit with just MT1 and MT2 swapped. The triac never fired and the bulb was never on for 15K, 33K, 100K, 200K

My head is now spinning. Can't really conclude here.

 

[ccurtis]

I looked at the data sheet for the BTA16-600B and it requires a whooping 25 mA of gate trigger current for quadrants 1-3 and 50 mA for quadrant 4. I think the difference in gate current for quadrant 4 explains why flipping MT1 and MT2 matters. The fourth quadrant is the quadrant where the gate voltage is positive and MT2 voltage is negative. The high gate current also explains why your resistors are hot.

I think you need a more sensitive triac for this application, say one with a max 5 mA gate trigger current like the BTA316-600D and connect the capacitor between MT1 and the gate. Alternatively, you can use a sensitive gate triac to trigger a large current triac with a not so sensitive gate.

 

[ccurtis]

The sketch requires a small correction, see attach.

I agree with the modification provided by EM2006. This will greatly reduce the power dissipation in the resistors. That is because when the triac turns on, the voltage across the circuit components drops to a very low level, waiting for the next half cycle. During triac turn-on almost all of the line voltage is dropped across the load (lamp).

Of course, at full diming the triac doesn't turn on, so R1 and R2 have to selected for full voltage, but the gate resistors only have to selected for the RC current at the mains voltage, but without continuous gate current as in the unmodified circuit.