Socket from immersion?

Bruthal

DublinDilbert wrote: »

As your going non-isolated with the cap dropper circuit, you could just use a 1MOhm resistor for your zero cross detector :eek: Its done quite often in some designs from the far east. I prefer to use an opto to bring in the zero cross, then on the output side use drive an optical triac to drive the main triac.

For dimming an incandescent bulb, the zero crossing might not be needed? Or maybe the dim setting would not be stable with wave chopping at varying stages over time? Probably no problem with a simple heated filiament though.
So with the non isolated supply to the circuit, maybe no opto is needed anyway. Could probably use a miniature transformer instead anyway which might be better, and go the opto isolator route then. Its all an experiment at this stage.

I remember about 15 years ago writing the assembly software for a college project, it was a triac based variable speed drive for a motor. I had run the software and it worked fine with a signal generator as the zero cross input. But when we hooked up all the hardware in the lab the thing would just go crazy. After about 10 mins I coped that the lab had more than one phase, the zero cross was on one phase and the load on another. Plugged the zero cross out of one socked and into the same as the load and it ran away perfect!

Id say you must of been there wondering what the hell is happening here, for a few seconds anyway. Thats an interesting one alright.

DublinDilbert

robbie7730 wrote: »

For dimming an incandescent bulb, the zero crossing might not be needed? Or maybe the dim setting would not be stable with wave chopping at varying stages over time? Probably no problem with a simple heated filiament though.
So with the non isolated supply to the circuit, maybe no opto is needed anyway. Could probably use a miniature transformer instead anyway which might be better, and go the opto isolator route then. Its all an experiment at this stage.
Id say you must of been there wondering what the hell is happening here, for a few second anyway. Thats an interesting one alright.

You'll really need to know the zero-crossing if you want to dim the bulb, cause once the triac is turned on it will conduct till the end of that cycle.

If your compiler doesn't support interrupts you'll have to poll the zero cross input, then do a delay loop of x cycles, then fire the triac. The larger the value of x less of the waveform that will be seen by the bulb. I can't remember the 12F pic off hand, but i think you have Timer0 and maybe Timer1.

If your compiler supports interrupts, just set up an interrupt on one of the pins I/O pins and an interrupt on Timer0 overflow. When you get your zero cross interrupt load Timer0 with your delay value. Then when timer 0 over-flows, send a short pulse output to turn on the traic.

If you want to mess around with this on the bench, just take a 50Hz waveform off the secondary of a small trafo and use a couple of resistors and you have a zero cross circuit.

For large heating/cooking loads they don't like delaying and firing the triac, they typically do a PWM type arrangement, where they will turn the load on for X of the 50 mains cycles per second. So if its on for 1 cycle out of 50, you'll get 1/50th. of the power. This will give a much better power factor for large loads. But they still measure zero cross so they know when to fire the triac.

Bruthal

Yea i get what your saying alright, using a microcontroller for dimming, it will need to know when to fire the triac with reference to the zero point of the cycle every time it fires the triac to conduct.