Can a TRIAC vary AC voltage?

[miceuz]

Is there an easy way to see if there is another winding on the transformer by checking the resistance between a couple of points?  If so, what should I measure?

You just have to have a look at the board, no need to measure - you will clearly see the primary winding and one or two secondaries. I can see 4 diodes arranged in full bridge configuration - those clearly are connected to one of the secondary winding which powers 7905. There is another diode, maybe that one is used as half-wave rectifier from another winding.

TL;DR; if there are 4 contacts on the secondary side of trafo, you have two secondary windings. If you have 3 connections on secondary side, you have a tranformer with middle-tapped secondary and that will be no good to power arduino as one side of tap is already "poisoned" by connecting to it mains.

[miceuz]

ok, I had a look at your photos myself, it seems there is only one secondary in the transformer. so, you need to power your arduino separately.

[scott216]

To power the Arduino could I just use an isolated DC-DC converter, something like this (Digikey 945-1655-5-ND):
http://www.digikey.com/product-detail/en/ROE-0505S/945-1655-5-ND/3461632

[scott216]

I took a shot at drawing a schematic of the toaster's circuitry.  See attached.

[Everton]

There is an error on your schematic because you have ground tied to fuse.  This will obviously blow the fuse every time.

[scott216]

There is an error on your schematic because you have ground tied to fuse.  This will obviously blow the fuse every time.

I'm petty sure the schematic is correct.  Ground in this schematic is not connected to an earth ground or anything.  If it was you'd be right. 

[David_AVD]

There is an error on your schematic because you have ground tied to fuse.  This will obviously blow the fuse every time.

In this case "GND" is simple the circuit's 0V reference.

[scott216]

I changed the circuit so Vcc is not a 5 volt offset from mains anymore (just had to remove one jumper).  I also removed diodes and resistors that tapped off the high side of the transformer.  I added an Arduino and a couple opto-couplers.  The Arduino works fine, but unfortunately, when the opto-coupler turns on, the triac does not turn on.  Attached is my revised circuit. 
In the original circuit, if I applied -5v to the triac (actually to the 91 ohm resistor in front of the triac), it would turn on.  But with the revisions I made that doesn't work anymore.  I even bypassed the opto-coupler and put the -5v on the 91 ohm resistor, but that didn't work either.
Any suggestions as to what's wrong?

[madires]

Reversed MT1 and MT2?

[scott216]

Reversed MT1 and MT2?

Are you asking if I tried reversing MT1 and MT2 on the triacs?  I have not.  I don't think it matters, does it?

[miceuz]

Any suggestions as to what's wrong?

Current can't flow via tracs gate as -5V is not mains referenced.

[madires]

Are you asking if I tried reversing MT1 and MT2 on the triacs?  I have not.  I don't think it matters, does it?

Please cross check with the original circuit diagram! Triacs don't work exaclty the same in all 4 operation quadrants. In Q4 the triggering is least sensitive. Some triacs won't trigger in Q4 at all.

[scott216]

So do I need something like this circuit:
http://www.flickr.com/photos/38917611@N08/4886233493/#sizes/l

I don't have a zero-crossover optocoupler triac like in the schematic above, but I do have an APT1221 phototriac coupler (data sheet).   Would that work?

[madires]

Just have a look at the original design. Is MT1 or MT2 connected to 120V?

[David_AVD]

Are you asking if I tried reversing MT1 and MT2 on the triacs?  I have not.  I don't think it matters, does it?

Please cross check with the original circuit diagram! Triacs don't work exaclty the same in all 4 operation quadrants. In Q4 the triggering is least sensitive. Some triacs won't trigger in Q4 at all.

That's exactly why I bought up the quadrant issue early on.  I suspect the OP didn't pay attention to it though.   

[scott216]

Just have a look at the original design. Is MT1 or MT2 connected to 120V?

MT1 (aka A1 on data sheet) is connected to 120 volt

[scott216]

That's exactly why I bought up the quadrant issue early on.  I suspect the OP didn't pay attention to it though.   

The whole quadrant aspect of triacs is new to me and I didn't really grasp the importance of it.  So I need to find some info on it.  I took a quick look at wikipedia and it seems like they have some good info.  If you know of any other good sources for a beginner please pass them on.

[madires]

There's http://www.onsemi.com/pub/Collateral/HBD855-D.PDF or http://www.nxp.com/documents/application_note/APPCHP6.pdf.

[Fank1]

A triac is nothing more than 2 SCR's back to back with a common gate.
You don't control the amount of current flowing thru them you only control when it turns on.
Each SCR turns off when the phase of the ac supply reverses.
You could control the on time and the off time to control the average power.
It's only a switch. ON or OFF.

[madires]

A triac is nothing more than 2 SCR's back to back with a common gate.

That's just a simple model to explain the basic idea behind it. The reality is more complex, as asual. A TRIAC can be triggered by a negative gate voltage. Can you do that with a SCR?

[max666]

Current can't flow via tracs gate as -5V is not mains referenced.

I agree with miceuz, by isolating the -5V from mains you also made it unusable as trigger voltage.

I think there is a crude way to return from pin 3 on the opto-couplers to neutral/mains with the proper resistance and somehow get the right trigger voltage/current. And there are also refined circuits with a capacitor and a Diac to actually trigger with a proper fast rising trigger current (as it should be), but I'm no expert on hot to properly trigger a Triac, so I would have actually hoped someone here would explain the Triac triggering 101 (or refer some read up).   

[scott216]

Ok, so I replaced my LTV-816 optocoupler with a APT1221 phototriac I had on hand.  I wired it between the triac and neutral - see attached schematic.  I found an example circuit using an Arduino to control a triac and I used this schematic as a reference.  Well, it blew the 91 ohm resistor in front of the triac.  I was worried about this before I powered it up because I didn't know what the current would be going into this resistor, I though that maybe the gate would limit it.  Also, I found a couple similar circuits like the one in the above example so I figured I'd give it a try.  My theory to this not working is that my phototriac is not a zero crossing, so the full 120 volts is going into the resistor.  Compared to a zero-crossing phototriac that's only on when the voltage is around zero and thus the current through the resistor is low.  Am I right about that?

Could I just replace the 91 ohm resistor with a bigger one?  If I had a 2400 ohm, then the current would be 50mA at 120 volts.  But this would be 6 watts of power, so I'd need a big resistor.  I don't really want to do that, but I want to know if that would theoretically work.

If this circuit will work with a zero-crossing phototriac instead of the APT1221, then that seems like the way to go.

[David_AVD]

Read up on the MOC302x and MOC304x series opto app notes.

[miceuz]

Well, it blew the 91 ohm resistor in front of the triac. 

Now your optotriac might be dead and the triac too. Maybe not, but have this in mind.

You have to limit the current thru the optotriac to account for the optotriac max current, but keep in mind the minimum latching current for the main triac. I've seen some advices on the net to make it worst case absolute maximum for optotriac. Usually, when everything is working, optotriac sees this current for a very brief moment - until the main triac is on, then the main triac shorts it thru itself. Plase note, that you have to calculate peak current for peak-to-peak mains voltage - 120V * sqrt(2) = 170V.

You have correctly noticed taht when using optotirac with ZC detection you would not have this problem as current would flow thru the resistor and opto until the main triac latches, you just would have to keep the opto ON for long enought for the current thru the gate to reach the latching current. Note that in this case you loose the possibility to do phase control.

[scott216]

Now your optotriac might be dead and the triac too.

optotriac still works, I haven't tested the triac.

Note that in this case you loose the possibility to do phase control.

I'm not going to do phase control at this point, just on-off.