I will study the document closely. Thanks an absolute bunch!

One last question if I may; how did your H11L2 hold up in this configuration? Wasn't it a little slow for a 100kHz signal?

 

I will study the document closely. Thanks an absolute bunch!

One last question if I may; how did your H11L2 hold up in this configuration? Wasn't it a little slow for a 100kHz signal?

I did not know your signal would be a 100kHz one ... that's pretty high ... is there a particular reason why you would want to run it that fast? That would require to carefully choose a dampening resistor at their gates (usually in the range of 10 to 100 ohms) plus an inverse parallel diode connected to the resistor to maximize discharge speed, among other things.

 

100kHz just to maximise the efficiency of the circuit. MOSFET's prefer higher frequencies, at least to my knowledge. Should I operate at a lower frequency? What frequency would you recommend so as to not run into your previously mentioned problems?

 

In the past, I've used the 6N137 opto isolator, instead of the H11L2. It's a bit faster.

 

100kHz just to maximise the efficiency of the circuit. MOSFET's prefer higher frequencies, at least to my knowledge. Should I operate at a lower frequency? What frequency would you recommend so as to not run into your previously mentioned problems?

Again, I need to know the nature of your application. Is it a motor? An incandescent lamp? A piezo component? ??

 

An incandescent lamp (50W) OR an LED lamp (7W). The whole idea is to make a dimmer for the lamp. It should work for the incandescent load and the LED load. Both lamps are rated to 230VAC

 

Does the LED lamp say that it is suitable for dimming?

 

No it doesn't. But I can always get lamps that are suitable for dimming. Let's just work off the incandescent for now.

 

At this voltage and current there is a good chance that the power dissipated as the MOSFETs turn on and off will be the main cause of power dissipation. Slowing down the PWM frequency is one way to reduce the power dissipation and therefore the heatsinking requirements. You may have a chance to test that concept soon enough

 

No it doesn't. But I can always get lamps that are suitable for dimming. Let's just work off the incandescent for now.

An incandescent lamp does not need that high a frequency to work well, at all. I've used frequencies of between 5 and 30 KHz in most of my projects, and I'd say that an incandescent of the power you've described will be more than well served with a frequency of between 10 and 15 KHz.

Word of advice, you need to use proper filtering at your MCU's power supply, otherwise very annoying spurious resets might happen due to noise generated by the PWM. Trust me, I know.

And another important thing, if I were you, I'd add a TVS diode in parallel to the load (in this case, the light bulb), so as to suppress any accidental spikes due to loose contacts or unintended disconnection while working. Normally a resistive load does not need a TVS, but I always place one as protection, just in case.

Since you've mentioned you will be working with 230VAC, it's peak to peak voltage is 230*sqrt(2)= 325V. In this case, I'd choose a TVS working at 400V, which is safely above the maximum working voltage, and well below the maximum 550V that the STP12NM50FP NFets are capable of handling. A good TVS to consider would be the 1.5KE440CA.