# Induction heater control --- simulation problem

Discussion in 'General Electronics Discussion' started by rickywangliwei, Jan 18, 2018.

1. ### rickywangliwei

3
0
Jan 18, 2018
Hi
I am planning to build a induction heater equipped with half bridge inverter and series rlc resonance structure. The power supply would come from line voltage and is then recitified with full bridge rectifier to supply the rlc tank. As the line voltage after rectification is 100Hz, I have to build the pll (the control loop) REALY fast. Before any hardware being made I decided to do some simulation first and the question came from the simulation model.

The model is relatively simple and it is essentially a pll: a vco drives igbt which is connected with the rlc tank. The current is fed back with an adc. The adc is used for online parameter estimation to produce actual load parameters like L and R and also for the phase detection.

In order to model the control loop the state variable representation of the plant was established and my question came from the phase detection and error signal. The state variables chosen were resistor voltage and the voltage across the inductor and capacitor in series. The reference was those two voltages with desired phase angle. Now what Im going to do is to compare the referece and the estimated state variables to produce the REAL TIME phase error so that it can be used to adjust the input of vco(actually a control block is needed to stablized the loop in between the error signal and the input of vco). After investing online I cann't find a easy way of comparing these two state trajectories and generate a error signal based on the phase difference of the reference and the feedback. Im now questioning if I was doing was right or not. I came to this point because I didn't want to use xor type phase detection because that would introduce delay and is hard to extract a precise model.

I now come to an dead end and really need your help. Anything concerning modelling and state variable feedback system would be very much appreciated!

BTW, the control loop parameter will be adjusted online(adaptive) according to estimated L and R so that the control loop is FAST and stable. The power control uses the frequency shift scheme.

cheers!

Last edited: Jan 18, 2018
2. ### kellys_eye

4,276
1,146
Jun 25, 2010
Why is 'fast' a requirement given the 'slow' response of the heating process? What do you hope to achieve by making the PLL lock/track faster? Just wondering....?

3. ### rickywangliwei

3
0
Jan 18, 2018
hi kellys_eye
its mainly because it has to keep up with the change of line voltage，otherwise the line current does not appear to be a sine， failing the emc
i hope its correct

4. ### Alec_t

2,788
742
Jul 7, 2015
Are you not smoothing the rectified mains?
You would never be able to keep up with all the changes due to spikes and other short-term variations caused by interference and load dumps.

5. ### rickywangliwei

3
0
Jan 18, 2018
Hi
a 2uf smoothing cap is in between the rectifier and the inverter

6. ### kellys_eye

4,276
1,146
Jun 25, 2010
Since the input side is 'SMPS-like' surely you'd be using the appropriate line filters, suppression components, rectification and smoothing of the AC input before application to the rest of the circuit.