Connect with us

PID Control of an Induction Cooker.

Discussion in 'General Electronics Discussion' started by Endre Furnes, Sep 27, 2014.

Scroll to continue with content
  1. Endre Furnes

    Endre Furnes

    15
    0
    Sep 24, 2014
    Dear all.

    I am a hobbyist in need for some advise. First a little bit background and description of my project. I am in need for a very precise temperature control of a large (50L) pot of water. Up till now I have done that with a 2kW ordinary cook top, a PID Controller (REX C100) and a SSR. Works fine, but 2kW is a bit sparse so I have bought an 3.5 kW induction cooker.

    The Induction Cooker has a rotary control (potentiometer) with and on / off switch integrated. 4 leaders to the pot, 0V - 5V - center tap and an on / off signal enabling the device.

    My take on this, in order to not doing any permanent and physical changes to the cooker, I just disconnect the flat control cable from the mainboard and set in on an 4PDT toggle switch and by that throwing the entire control out to external. By that I can toggle between ordinary and external control.

    The PID Controller (REX C100) has a 5V PWM signal, periods adjustable but min. 1 sec.

    An second order low pass RC filter (C1-R3-C2-R4) deals with that without too much ripple on the voltage. Somewhat slow in response, but no concern as after all it is 50L of water to change :)

    I have not build the circuit yet, as I am still waiting for some parts, but simulates in LTSpice.

    First 'problem'. By opening and measuring the voltage on the control knob, I discovered it is not a 0 - 5V, but rather a 5 - 0V signal. The center tap rests against the 5V side of the pot. at minimum setting. I needed to inverse the signal, and found that best to do before the filter. I uses a transistor (Q1) as a NOT gate, and that seems to be working fine as well. I have access to both 0V and 5V from the cooker, and I uses that on the transistor, no need to build an voltage supply.

    Second problem, and here I do need help. It turns out that the minimum power on the induction cooker is 500W. It goes to 500W immediately after enabling the device. Hmmm ... after testing, 500W is enough to slowly, slowly rise the temperature even in this large amount of water, so I do need to implement the on/off function.

    What I was hoping to achieve was an accumulation of voltage in C3 from the PWM signal, so that if no pulses had been on for a while, C3 would drain and cutting the on / off line switching the device off. By the first puls from the PID, C3 would charge and opening the line again.

    My smoothed signal would be better to use, but that is not so easy, since it is inverse.

    I do need some advise her, gentlemen :)

    Regards

    Endre
     

    Attached Files:

  2. Endre Furnes

    Endre Furnes

    15
    0
    Sep 24, 2014
    I have seen it before, - writing up clears one's mind.

    As I write 'my smoothed signal would be better to use', - a parallel filter before the NOT gate gives me a decent 0 - 5 V to handle. But how is it best to switch it?
     

    Attached Files:

  3. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    Hi Endre
    What PWM timings can you output from the PID i.e. on and off timings.
    Adam
     
  4. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    At what PWM timings would you like it to turn off.
    Adam
     
  5. Endre Furnes

    Endre Furnes

    15
    0
    Sep 24, 2014
    Hi Adam. Thks for the interest.

    Well, - the timing is adjustable, but the shortest period i can set is 1 sec. To turn off the line, I am hoping for a delay ... that the pwm has been fully off for a 'while', some 15 sec or around there.

    At the first puls of signal, the line back on.
     
  6. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    When you say PWM fully off do you mean it is switched off?
    Adam
     
  7. Endre Furnes

    Endre Furnes

    15
    0
    Sep 24, 2014
    No, not switched off, - just a 0 - signal. No 5V pulses, - just a steady 0V.

    Endre
     
  8. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    I thought your minimum period was 1 second so how can you have a steady 0V?
    adam
     
  9. Endre Furnes

    Endre Furnes

    15
    0
    Sep 24, 2014
    Hi again Adam. I'm not sure I use the right words. With 'period' I mean the portion of time the pwm is 'chopping' up.

    At 50% output, the signal is 5V 0.5 sec and 0V 0.5 sec. At 10% output the signal is 5V 0.1 sec and 0V 0.9 sec.

    When the PID output goes to 0, - there are no pulses, so a steady 0V.

    Endre
     
  10. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    This is what I mean, both these signals have a period of 1 second but different on times.
    Adam
    PWM.jpg
     
  11. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    No I think your right I just needed the answer you just posted.
    Adam
     
  12. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    Ok so when do you want it off, when the signal is off for 0.9 seconds?
    Adam
     
  13. Endre Furnes

    Endre Furnes

    15
    0
    Sep 24, 2014
    I hoped to switch off the line after there has been no output from the PID for some sec, - say 15 sec.

    Maybe I need a timer for this?

    Endre
     
  14. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    He is a circuit that I think might work. You may have to change the timing components for what you want. But it shows the circuit switching off after about 10 seconds and then on again when it sees the next pulse. If it sees a pulse within this time the circuit doesn't switch off. You will need to add some supply decoupling capacitors which I have not shown.
    Adam

    PWNCCT.JPG

    PWM.jpg
     
  15. Endre Furnes

    Endre Furnes

    15
    0
    Sep 24, 2014
    This is great Adam. Thank you so much. I do have to study this in order to understand, and with my slow mind it might take most of the evening :)

    I am a complete rookie when it comes to circuit engineering, as I used to be a mechanical engineer before retirement.

    Again thanks.

    Endre
     
  16. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    I am also slow, I make as much progress sometimes as an asthmatic ant carrying heavy shopping. But when I do understand things it is usually there to stay.
    All the best Adam
     
  17. Endre Furnes

    Endre Furnes

    15
    0
    Sep 24, 2014
    Well Adam :)

    I have now implemented your suggestion to my circuit and it seems to be just what I hoped for.

    Timing components that would be the capacitance of your C1 and the resistance value of your R1 I assume? Yes, timing will be essential, as most of the time the controller will work between 0 and 500W to keep an steady temp.

    By the way, - do you mind enlighten me around your statement 'You will need to add some supply decoupling capacitors which I have not shown.' as I don't know what that is?

    I owe you one, Adam

    Regards
    Endre
     

    Attached Files:

  18. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    You dont owe me anything, it's my pleasure. So nice to deal with such a greatfull person for a change.
    Adam
     
  19. Arouse1973

    Arouse1973 Adam

    5,164
    1,081
    Dec 18, 2013
    Most circuit will require some capacitance across the supply to ensure stability. I would put a 22uF and a 100nF across the supply lines for your circuit. Make sure the voltage rating of the capacitors is above the max D.C voltage of your circuit

    http://en.m.wikipedia.org/wiki/Decoupling_capacitor
    Adam
     
  20. Endre Furnes

    Endre Furnes

    15
    0
    Sep 24, 2014
    Yes, yes ... I see what you mean now even before reading your link. There is 19 components on the board, with various needs at various times, sucking juice from tiny wires.

    It is fascinating to see the similarities with pneumatic and hydraulic as I used to work with in the 70s and 80s.

    Thanks again. Yes, I owe you one

    Endre
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-