help with small digital induction heater circuit. PIC, ZVS, ZCS, MOSFET or IGBT

[fastline]

playing around with a simple circuit for relatively low power and lower voltage heating. I was originally playing with a self oscillating circuit but decided I was living and playing in the past and wanted to work this from a modern level.

Due to the power levels, I am really trying to do this with just a bridge rect. From what I read, chasing some ZVS/ZCS techniques will help minimize losses, heat, etc. I am still unsure if I should just use MOSFETs in the design or use IGBTs as they are more recommended for zero switching??

I want to vary the voltage at the work coil and not yet sure of a cheap way to do this other than a SMPS? Again, trying to stay away from that right now. Yes, I can vary all this with a DC lab supply but I want an operating circuit from 120VAC.

One of the reasons I wanted to work on digital was I found all the expensive capacitors prohibitive. They were going to get very hot, needed cooling, and seemed I was chasing old school tech.

Can someone help me get up to speed>

 

[fastline]

I am finding that I need to better school myself to play in this area but also finding not many people know enough, like me.

I was originally playing with a hobby circuit using a center tapped work coil which is a nice little self oscillating circuit. However, there are two things I was trying to remedy by trying to improve by implementing PIC control of the circuit is A) reduce the need of high quality, low ESR, high value caps which will still heat up and cause cooling problems. B) Improve the efficiency of the overall design with will also improve heating problems.

It became obvious that switching losses can be an issue here so zero switching is needed. IIUC, you would still need to tune the exact same circuit's oscillating frequency and program the MCU accordingly to switch at near zero switching. Only difference I can see right now is the MCU would have tight control of actual frequency and other safety functions built in such as thermistor inputs, etc. It just becomes a PLC?

I am just trying to design a simple, efficient induction circuit. I would tune the optimal frequency via R&D and use only one frequency.

Can I reduce my parts list and caps requirement using an MCU?

 

[moffy]

A MOSFET or IGBT would be fine. But what is the power rating of the heater?

 

[DerStrom8]

Hello fastline,

Seeing as both threads appeared to be for the same project, I have combined them. It's generally better to keep one thread per topic

Regards,
Matt

 

[fastline]

I am trying to target power at around 1000w but for the R&D circuit, I am pretty open there.

I originally wanted to apply a modern MCU to this project but I guess I am now second guessing if that can really be beneficial to efficiency and parts expense.

 

[moffy]

You could just use a triac as the switching/phase control element. Use a MOC3020 - MOC3023 as the opto isolated control element, and use a MCU to control the MOC302x.

 

[fastline]

Thank you. Right now I think my questions are very specific as to "why" or "to what benefit" the MCU plays in this circuit? As well, I am trying to determine some benchmarks to determine how much capacitor heating I might see. I was recently reviewing a modern cooktop design that just had caps mounted to the PCB and nothing more!! Then you see DIY guys with vids of smoking caps, water cooling, etc. I need to look closer at this stuff.

 

[moffy]

Sorry, I missed the "Induction" bit. You will need a resonant L(being the induction coil) C circuit, driven probably by MOSFETS because of the high frequency.