PLL For Induction Heating

[Alexsgarage]

What would be the best way to create a phase locked loop for an induction heater? I will use a XOR logic gate as the phase detector. The inputs will be the inverter and tank voltages (clipped by diodes of course) turned to square waves with a voltage comparator. How would I set up the output of the XOR gate to the oscillator?

 

[marcbarker]

What would be the best way to create a phase locked loop for an induction heater? I will use a XOR logic gate as the phase detector. The inputs will be the inverter and tank voltages (clipped by diodes of course) turned to square waves with a voltage comparator. How would I set up the output of the XOR gate to the oscillator?

I'm very curious the reason for phase-locking an induction heater to anything else. Are you wishing to synchronise it to something? I would of thought randomly synch'd IH were better for EMC.

So curious I am, that I'm prepared to help you design it.

 

[Alexsgarage]

As the tank coil will change inductance due to different workpieces, the resonant frequency of the LC circuit will change. If the LC resonant circuit can be locked to the inverter frequency then the device will automatically asjust the oscillator to accomidate different resonant frequencies of the tank.

 

[marcbarker]

As the tank coil will change inductance due to different workpieces, the resonant frequency of the LC circuit will change. If the LC resonant circuit can be locked to the inverter frequency then the device will automatically asjust the oscillator to accomidate different resonant frequencies of the tank.

Sounds to me like you're making a lot of unneccesary extra work for yourself. You don't need a PLL to do that, but if you really want to do it this way, I'll still help

 

[Alexsgarage]

What would be the best way to do this then? I still want the oscillator to adjust to the resonant frequency automatically for maximum efficency. How would be the best way to do this?

 

[marcbarker]

What would be the best way to do this then? I still want the oscillator to adjust to the resonant frequency automatically for maximum efficency. How would be the best way to do this?

The easiest way I think is utilise the laws of physics to your advantage:

"In an oscillator circuit, the overall frequency of oscillation is dependent on the resonant circuit having the highest Q"

Or in other words, just make the whole thing self-oscillating and then 'let go' of it. When it's running, it's a "Class-C oscillator". Funnily enough Class C oscillators are very commonly used in Induction Heating.

Even more funnily enough, a recent thread I was posting to recently was about a Class C oscillator used in a transmitter https://www.electro-tech-online.com/threads/uni-project-fm-trnasmitter.95651/

Sorry, but I say chuck away ideas about PLLs and that , but if you still want to do a PLL I'll assist.

 

[Alexsgarage]

Well I have 4 IGBTs coming and I have spent a quite a while designing drivers, oscillators and power supplies. I also designed an oscillator circuit using an SG3524 PWM and all of this took time and money to get to so that being said I dont really want to change my plan at this point, I will post all of my schematics in a bit when I get my laptop back and then we can work from there?

 

[marcbarker]

I understand where you're coming from and recognise the considerable investment in both time and money that you'd made so far.

Please post what you have so far and I'll see what I can do for a PLL-based solution.

Even if it's 10x more complicated than it needs to be, it is a golden learning experience

 

[Alexsgarage]

Ok, cool, thanks, I will post the schematics soon.

 

[Alexsgarage]

Here is the oscillator schematic. It uses an SG3524 PWM, the power supply connections to the right of the chip are for the floating supply as not to short out the output. R1 and C2 adjust the frequency.

 

[marcbarker]

OK, this is going to be the 'VCO' right? So R1 will need to be a BJT as a current sink, it's base becomes the VCO control voltage.

What's next on the agenda then?

Phase comparator? Loop low pass filter? Crystal controlled reference frequency generator to synchronise this VCO to?

 

[Mr RB]

PLL??

No no no... He needs a PID controlled PLL with a Heisenberg compensator.

 

[Alexsgarage]

Ok, I will change the schematic to add the BJT in place of R1. Also I will use a XOR logic gate as my phase detector, that will be posted soon.

 

[Alexsgarage]

Here is my phase detector, how can I connect the output of this to my VCO?

 

[marcbarker]

In an analogue PLL, the phase comparator feeds the VCO via a LowPassFilter. It's an L-Pad. It's capacitor has a resistor in series.

You're suppose to calculate the LPF, and there's a fair bit of maths, but you can tune 'by ear' instead with a scope if make the R's variable.

So, what you going to synchronise the VFO to?

And, Is there a block diagram you can post yet? (kinda important!)



edit: I believe you're going to use the instantaneous phase of the output LC as the other input to the PC.

If so I assume the object of the PLL is to null the phase lead/lag between the (VCO) driven IGBT (active) phase and the LC (passive) circuit phase.

 

[Alexsgarage]

Here is the block diagram.

 

[Alexsgarage]

Yes, that is exactly what I want to do.

 

[marcbarker]

Ok what's the next step then? Can I assist?

 

[Alexsgarage]

Yes, I dont really know the best way to design the charge pump, do you know of any schematics?

 

[Alexsgarage]

An update on the SG3524 oscillator circuit.