# Tank resonance locator

Discussion in 'Electronic Projects Design/Ideas/Reviews' started by Vijit Dubey, Sep 21, 2015.

1. ### Vijit DubeyMember

Joined:
Sep 21, 2015
Messages:
56
Likes:
0
Location:
College Station, Texas
Let's say I have a capacitor and inductor (used in a tank circuit for induction heating application). It operates at a PLL using CD4046. However, to calibrate the starting frequency I must know the tank's resonant frequency. Using either a micro-controller or any other logic, what would be the best way to identify the tank's resonance frequency by sweeping from a high frequency in such a way that I get a number displayed as the resonance of the circuit.

2. ### alec_tWell-Known MemberMost Helpful Member

Joined:
Jul 10, 2011
Messages:
9,318
Likes:
1,230
Location:
Cardiff, Wales
Welcome to ETO!
You could put a current-sense resistor in series with the tank circuit and check the voltage developed across the resistor. For a parallel LC tank the current will be a minimum at resonance. Display the frequency corresponding to that minimum.

• Like x 1
3. ### Vijit DubeyMember

Joined:
Sep 21, 2015
Messages:
56
Likes:
0
Location:
College Station, Texas
It is a series LCR. R being the load in the induction heating. Now, lets say L and C are like a blackbox and I also have a current feedback too.

How can I develop the following logic in a new test circuit that I will build:
• Frequency starts at a high value, it keeps on decreasing and senses the current and voltage. The moment circuit reaches the maximum current of frequency it stops and returns me a value of the frequency.
I want this circuit to be a test circuit each time a customer may get a new LC so that time is not wasted. What IC and logic could I use? I tried looking a lot of papers but could not find one. I know it should not be very complex.

Vijit.

Joined:
Jan 12, 1997
Messages:
-
Likes:
0

5. ### AnalogKidWell-Known Member

Joined:
Sep 22, 2011
Messages:
1,135
Likes:
161
Location:
Ohio - GMT-5

Probably not, but too many possible ways. A block diagram of your setup will decrease greatly the verbal back and forth to work out the details of the question.

ak

6. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,128
Likes:
564
Location:
AZ 86334
How about pulsing the tank, and measure the ringing frequency? Is the Q of the tank too low to ring?

7. ### Tony StewartWell-Known MemberMost Helpful Member

Joined:
Aug 31, 2012
Messages:
3,152
Likes:
281
Location:
Richmond Hill , ON Canada near Toronto
This can be done in software or hardware in either amplitude peaking with a rectifier measurement circuit or a phase as resonance using the MC4046 phase detector.

This freq. will lower with a various pots on stove top due to permeability but also lower in Q due to power loading.

If I were doing it, I might consider other methods.

Similar to Solar tracking which has a low Q Peak Power tracking point, a loaded induction coil will have low Q.
So if you wish maximum power, depending on freq and if high f; alum. or iron pot the loaded Q may not affect power significantly or not.

Solar tracking max power regulators use two MPPT methods worth considering perturbation and observation, (P&O) and InCon or Incremental Conductance

Solar uses DC , in this case AC so measuring phase current relative to voltage resolves optimum frequency, but you need to monitor load power in case of no load to shut off or monitor self heating or OTP or OCP from a no load situation.

So I would sweep f and measure power for peak and no load protection in addition to power modulation with power sensing. (V*I ) if you wanted to regulate power with dial regardless of load.

• Like x 1
8. ### Vijit DubeyMember

Joined:
Sep 21, 2015
Messages:
56
Likes:
0
Location:
College Station, Texas
Thank you for replying AnalogKid and MikeMl
I have come up with a little explanation in the attachment for what I am looking for. I want to make a independent circuit which conducts some tests and gives me the tank's resonant frequency as a number. I hope it is quite clear this time.
The application is for induction heating, where L and C have a very wide range. However, L is typically around 1uH, C is around 6-8 mF and it would be tested in no load condition (no iron job between coil), so I don't know what to comment on R. Now these L and C work for around 50-60 KHz.
But I want to create a test circuit which can tell me the resonant frequency of my tank from the range of 3KHz to 150KHz.

9. ### Vijit DubeyMember

Joined:
Sep 21, 2015
Messages:
56
Likes:
0
Location:
College Station, Texas
If I use 4046 and feed a current feedback to the pin 14 and compare it to VCO out in XOR comparison. What do I do with the output pin 2. What do I trigger with it. Also, I can't seem to understand how would this help me measure resonance current.
Here is a layout of what I am talking about.

4046 IC:

Vijit.

10. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,128
Likes:
564
Location:
AZ 86334
Feed V(Burst) to a ucontroller, and write code to measure the Period. The resonant freq of the tank is 1/Period.

• Like x 1
11. ### Vijit DubeyMember

Joined:
Sep 21, 2015
Messages:
56
Likes:
0
Location:
College Station, Texas
It looks good in simulation but I am not sure if I could do it in real life. Also, I don't have a ucontroller. I however do have access to a lot of ICs, and I am looking to make it using a IC preferably. Do you think I could get something out using CD4046.

12. ### MikeMlWell-Known MemberMost Helpful Member

Joined:
Mar 17, 2009
Messages:
11,128
Likes:
564
Location:
AZ 86334
Why do you even need a PLL? Just make the tank circuit part of a power oscillator. It will naturally oscillate at its resonant frequency.

• Like x 1
13. ### Vijit DubeyMember

Joined:
Sep 21, 2015
Messages:
56
Likes:
0
Location:
College Station, Texas
Okay, let's say I use a power oscillator. Can I use a VCO (voltage controlled oscillator). If yes can I use VCO in CD4046 (pins 4-12)
And feed the output of VCO to what?
And if circuit will resonate at its natural frequency, how can I measure the value of frequency.
Sorry for bad questions. I am a little confused now.

Vijit.

14. ### Tony StewartWell-Known MemberMost Helpful Member

Joined:
Aug 31, 2012
Messages:
3,152
Likes:
281
Location:
Richmond Hill , ON Canada near Toronto
The easy way is to use an LCR meter to read the L or C values at 100kHz in case one or the other is out. ( both are high tolerance)
The next easy way is use a DSO and pulse it and measure f on screen.

Failing that..... DIY

You can automate or manually sweep VCO to drive Type II PD so that voltage out to suitable filter indicates phase lead or phase lag where resonance gives phase for Voltage vs Current.

Use a suitable current Shunt R. and bias to V/2.

BTW I think low f Induction heaters are not as good as high f , which works on Alum pressure cookers etc.
So reconsider your L C and Q values.

15. ### Tony StewartWell-Known MemberMost Helpful Member

Joined:
Aug 31, 2012
Messages:
3,152
Likes:
281
Location:
Richmond Hill , ON Canada near Toronto
Disregard extra VCO AC coupler, clamp diode needed for my mickeymouse simulator to bias inputs., That's supposed to be a CD4046 Type II phase/Frequency mixer,

16. ### Vijit DubeyMember

Joined:
Sep 21, 2015
Messages:
56
Likes:
0
Location:
College Station, Texas
Tony Stewart This is too confusing. The thing is that I am making this project to present with a company, so I cannot negotiate with values.
Is there a way I observe the peak value of current (tweak from high freq to low) and stop the sweep the moment the amlitude starts decreasing. Please consider that I have to implement it directly, simulations won't help.

Vijit.

17. ### Vijit DubeyMember

Joined:
Sep 21, 2015
Messages:
56
Likes:
0
Location:
College Station, Texas

18. ### Tony StewartWell-Known MemberMost Helpful Member

Joined:
Aug 31, 2012
Messages:
3,152
Likes:
281
Location:
Richmond Hill , ON Canada near Toronto
This is a proof of concept that works. I showed a Series current shunt. It should be less than coil Rs so that Q or power is not affected in a real operation. But for measurement, of resonance it doesn't matter.
All you need is a pot & VCO, counter with Type II out and filter or a sig gen and scope or pulse gen and DSO

It depends on your limitations and equipment avail and the test process specs.

I can think of 10 ways to do this. But the best way matches your incomplete requirements above.and lowest cost.

19. ### Tony StewartWell-Known MemberMost Helpful Member

Joined:
Aug 31, 2012
Messages:
3,152
Likes:
281
Location:
Richmond Hill , ON Canada near Toronto
?? What equip. is avail? What is the overall objective? How to select? or tune? or show company how to test it?

20. ### Vijit DubeyMember

Joined:
Sep 21, 2015
Messages:
56
Likes:
0
Location:
College Station, Texas
Few IC (like CD4046), oscilloscope, other analog circuit elements are available.
The objective is to know the natural freq (number) of a LC together, so the eventual aim is to build a circuit which is robust and mobile and can tell me the natural freq in a few minutes (and even untrained person can use it)

Also, I want to use low current (since circuit is frequency dependent). Hence, it would be better with a voltage feedback.

21. ### Tony StewartWell-Known MemberMost Helpful Member

Joined:
Aug 31, 2012
Messages:
3,152
Likes:
281
Location:
Richmond Hill , ON Canada near Toronto
How will you calibrate frequency? Is it analog or DSO?
It should only take a few seconds to measure .

How accurate? Is a freq counter avail, or just for calibration?

Last edited: Sep 21, 2015