Mechanically controlling a InductionCooktop - Need help for alternate solution

[ngg]

Hello all

I am trying to control an Induction Cooktop using a Arduino. I am using this model https://www.flipkart.com/cello-blazing-600-induction-cooktop/p/itmeg3f99rptbje8?pid=ICTEG3F9KBWMA8QR&lid=LSTICTEG3F9KBWMA8QRT0FQUJ

What have I done so far is control the temperature inside the vessel on the cooktop by using a servo motor + Arduino to rotate the potentiometer knob (It's a B500K pot) and use a DS18B20 thermometer to get the feedback for the PID controller which maintains the temperature.

I have attached some pics to better explain my setup, if you guys need anything please let me know will add.

I am looking for advise in how I can control the induction cooktop digitally without a servo motor in between.And if possible get temperature reading from the thermistor in the center of the induction coil, and identify if there is a vessel is present or not.

Any help is much appreciated. I am an amateur so please be a little elaborate if you any suggestions.

Options I have in mind so far:
- Try to hack the UART communication - FAILED as I think there is high voltage floating, I get shock when I connect a logic analyzer, also I am not sure if its UART.
- Use opto-isolator + Digital potentiometer combination

[soldar]

The description page you linked to says the top work surface is steel. Is it really steel? I have never seen an induction cooktop made with steel.

Interesting also to see the gears that drive the pot. Did you make those yourself?

At any rate, I would first determine what kind of signal goes through the pot. If it is low voltage DC the solution is going to be easier than if it is high voltage AC. To start you can use a multimeter (with great care).

[ngg]

The top of the cooktop is ceramic glass. The picture I posted shows the internals.

Yes, I 3d printed the gears.

I think it has both AC & DC.

DC voltage varies from 0-3.5V when I connect the multimeter across the pot and turn it, and I get AC voltage of ~8.

[max_torque]

The screen printing on the connectors is ambiguous!  At one end it says "clk & Data" and at the other "TX & RX".. One suggests a serial protocol with a seperate clock (ie SPI etc) the other a protocol with an integral clock (ie UART).  In either case, the voltage levels are almost certainly not ground referenced because they float in the application.  The shock you get could be a DC offset or an AC perterbation.  To put a scope on those signals you'll either have to "float" your scope, or float the device itself.


As the signals look to be digital, rather than analogue, perhaps the pot that comes in the unit is not actually an analogue pot, or is there some sort of decoder IC we can't see on the small "pot interface" pcb?

[soldar]

The top of the cooktop is ceramic glass.

Yes, it would have to be.

Yes, I 3d printed the gears.

I am impressed. Good job.

I think it has both AC & DC.

DC voltage varies from 0-3.5V when I connect the multimeter across the pot and turn it, and I get AC voltage of ~8.

I think it might probably be some sawtooth or such shape that only has positive voltage in which case you can use a regular transistor or even optocoupler. You should find out with a scope.

[Nusa]

You could go hunting for an appropriate digital potentiometer IC, aka digipot.

Or you could go with another low tech solution, such as relays attached to fixed resistors to pick different power levels. You could use one relay for each power level you need, or select resistor values such that you can pick a range of power levels depending on which combination of relays are active.

Or the optoisolator/transistor equivalent.

The solution you already have will likely work too. You've created a motorized potentiometer, which can be found as a commercial product as well.

Or you can figure out how the control circuitry on the main board actually works and skip the child boards.

At one end it says "clk & Data"

The labeling on that end is clearly for a 5-pin connector, in which only a 4-pin connector is installed. So who knows if they apply logically or not. One would have to know more about the circuit as implemented to figure it out.