Triac sinking 30VAC, switched with 12VDC?

Thread Starter

JC37

Joined Nov 5, 2016
1
So this is sort of an odd series of questions. I'm currently working on some boards that use a 555 timer running on 12Vdc to switch the ground side of a 30Vac light bulb, using a 2N6075A triac. The 12VDC is produced by running the 30Vac through a single diode to produce 30Vdc, stabilizing it with a 680uf capacitor, and then running it through a 12V regulator. The output pin of the 555 is connected to a 1k resistor, then the gate pin of the triac, then to a 10k resistor, and finally ground. The lightbulb is connected to AC supply and I believe MT2 of the triac, with MT1 grounded.

From what little I've learned about triacs (I don't see much AC), this *should* result in essentially only the positive side of the AC sine wave being passed through...and yet the whole thing seems to be passing through. If I use the 30Vac to trigger the gate pin, the brightness of the bulb doesn't change, and if I trigger it using the AC through a diode, either direction, the brightness drops. I haven't yet looked at it on a scope, but I'd say chances are good the whole wave is being passed using only DC. Is this something special about the 2N6075A? I can find nothing about DC switching capability in the datasheet, so I'm trying to understand *why* this works, when everything says it shouldn't.

One thing I am seeing is that at the gate pin of the triac, when it's driven high using DC voltage, at the gate pin I'm measuring 0.455Vdc and 0.560Vac. Is this AC voltage leakage through the gate pin, or just noise that my meter's not smart enough to filter out? If I go to the 12Vdc supply, the AC is completely gone.

But now, onto the real question. Because of the high rate of failure of the 555 timers and a few other design flaws, and also because the designer has passed away and replacement boards aren't available, I'd like to design a new board, replacing the 3 555s and a lot of the other timing components with a single PIC12F. Since these boards may end up in other locations where they're switching 12Vdc instead of 30Vac, and I'd like to make one board with a couple jumpers to cover everything, the PIC will be driving transistors anyways, probably 2N3904s. I'm wondering if I can safely switch those 2N6075A's using 12Vdc through the 2N3904's, or if the AC voltage I'm seeing at the gate will cause them to fail, and may be the reason the 555s are failing. Anyone have a guess?
 

AlbertHall

Joined Jun 4, 2014
12,345
Triacs can be triggered in all four quadrants - that is either polarity of gate voltage can trigger the triac for either polarity of voltage across the triac. So your 12V on the gate will pass the whole AC cycle to the lamp.

Now about the AC on the triac gate, it isn't the 30V AC coming back out of the gate. I suspect (I have never measured this) the necessary gate current will be different for each cycle of the AC and this varying current will result in different gate voltages. Your meter will read the average DC and will see the variation of voltage as AC.

Yes, you can use a PIC to drive transistors and the transistors can drive a triac given a suitable circuit arrangement.
 

Alec_t

Joined Sep 17, 2013
14,280
when it's driven high using DC voltage, at the gate pin I'm measuring 0.455Vdc and 0.560Vac. Is this AC voltage leakage through the gate pin, or just noise that my meter's not smart enough to filter out?
Your meter probably computes RMS AC voltage on the basis of an expected pure sine-wave input, so is being confused by the non-sinusoidal gate voltage which results from the gate current.
 

RichardO

Joined May 4, 2013
2,270
I you don't want the full ac to pass through then you can use an SCR in place of the triac. An SCR only passes current when the anode is more positive than the cathode.
 

crutschow

Joined Mar 14, 2008
34,281
Note that you can't use BJT transistors to switch AC, if that was what you were thinking.
Edit:
If you do want a switch for AC, or plus or minus DC using transistors, here's the LTspice simulation of a circuit using two back-to-back N-MOSFETs and one PNP BJT to form a low-side AC/DC bipolar switch.
This has much lower On voltage drop and power loss than a TRIAC and it doesn't latch on with a DC source as a TRIAC does.
The PNP allows the positive control voltage to turn on the MOSFETs, but blocks the negative voltages that occur when the control signal is low and the MOSFETs are off (Vgs at zero).
The input control voltage must be equal to the voltage (Vgs) for which the ON resistance, Rds(on), is specified in the MOSFET spec sheet.
This is typically 10V for standard MOSFETs and 5V or less for logic-level type MOSFETs.

upload_2016-11-6_23-10-36.png
 

Attachments

Last edited:
Top