Series or Parallel connection for LC circuit

[baconman71]

I'm trying to make a wireless energy experiment by using LC circuits as resonators. But within this circuit you can either make the inductor and capacitor in series or in parallel. What do the different combinations do to the circuit? Is there one that is ideal for some circumstances?

 

[jsgruszynski]

It depends on what kind of resonance you want:

For parallel, at resonance the impedance goes to "infinity" (or to the leakage R in parallel with the LC)

For series, at resonance the impedance goes to zero (or the parasitic R in series with the LC)

Thus is depends on what your circuit design and intent is.

If you look at a typical radio front-end, they use a parallel LC which means that at resonance (at the tuned frequency of the radio), the impedance looks large so you get the largest voltage drop at that frequency but at all other frequencies the voltage drop is approaching zero.

If you look at a typical antenna tuner for a similar radio, they use a series LC which means that the at resonance it looks like a short but at any other frequency it look like a large attenuating impedance.

 

[baconman71]

Ok, so from what I understand it would be better to put the two connections (on the sending and receiving coil) in series to allow for maximum power to be transmitted?

 

[NascentOxygen]

Ok, so from what I understand it would be better to put the two connections (on the sending and receiving coil) in series to allow for maximum power to be transmitted?

What circuit/device are you planning to connect to the "sending" resonator? What are you connecting to the "receiving" resonator?

Will the resonators be spaced just a few cm apart, to demonstrate that energy can be transferred?

 

[sophiecentaur]

You are presumably trying to connect a signal source to an antenna of some sort??
You need to do two things, you need to 'tune out' the reactive parts of the antenna (if it is not already resonant) and then you need to 'match' to the output impedance of the source (50Ohms?). What you do will depend entirely on the antenna type and size. There is no single answer. Look up "Antenna Matching" Images to get an idea of the range of possible designs.

 

[baconman71]

I am planning to probably attempt to try to power several different items beginning with an LED or maybe just using a multimeter to see how much power is actually transferred. And sophiecentaur what do you mean by match output impedance of the source? I'm just confused on that.

 

[sophiecentaur]

I should point out, from the start, that this business is potentially very complicated but that a bit of knowledge at the start can give you a good chance of getting better results for your experiment than you'll get by just 'piling in. The possible improvement in results could be impressive. The point of having a resonator is to 'match' the transmitting source to the antenna.
First, it is vital to have an idea of the sort of frequency you plan to use. That affects the optimum design of antenna and matching network. What is your RF source?

You may remember about battery internal resistance and that it causes a loss of PD when a current passes. As you lower the load resistance, the output voltage drops but the current goes up (as you'd expect). The Power delivered to the load is V times I and it is a maximum when the load resistance is equal to the battery internal resistance. At this point the battery will be getting very hot, of course (it is dissipating an equal amount of power inside itself) and would not be long for this world but the load is 'matched' to the source resistance.
Likewise, if you want to get the most power delivered into an antenna, you need to match it to the source. (This involves Reactance as well as resistance but the same principle applies as for a battery at DC) A matching network, consisting of an appropriate coil and, possibly, a capacitor, can achieve this at one frequency. The 'resonator' is a matching network and will help to maximise the power delivered. If you just hook up the 50Ohm output of a signal generator to a piece of wire, the load will just appear as a tiny capacitor (very high reactance) in series with a very low resistance (which is due to the energy radiated out into space). You won't be radiating much power at all. A series coil can resonate with the capacity of the wire and then you just end up with a low resistance - an improvement.

A more fruitful approach for low frequencies would almost certainly be to use a loop antenna, consisting of a large coil of several turns of wire on a frame (say 1m across). This should work better, on its own, than a wire and can still be resonated with a capacitor across the 'feed' terminals.

But it would be easier to help you more if we could know what you actually plan to do. (Remember that there are regulations governing what you are allowed to radiate in the RF bands!)

Google MF loop antenna matching youtube. There is a good selection of nerdy lads showing you what they have done.

 

[baconman71]

Ok let me describe it to you. I want to mimic the idea that MIT used on a project they did several years ago that has now turned into witricity. First there is the source of the frequency. I have been told that the "higher" the better and I have been told to try making an oscillator of sorts. Then from there I will calculate the LC circuits on both sending and receiving sides. This lc circuit was used by MIT and actually many other people trying the same system because it is supposed to help get even greater distance out of transferring power. So just an inductor and capacitor that are set to resonate at the frequency created by the oscillator. Then the receiving coil will resonate at the same frequency. I don't what frequency I want to use but I have been told that higher KHz and lower MHz would be fine.

 

[NascentOxygen]

I think your first step should be to find an oscillator circuit of roughly that frequency and build it. Then investigate its frequency and waveform. Do you have an oscilloscope?

 

[baconman71]

Ok good. I am actually in this stage right now and I do thankfully have an oscilloscope or else this would be infinitely harder. What would be the step after this?

 

[baconman71]

http://digitalcommons.calpoly.edu/c...ontext=eesp&sei-redir=1&referer=http://schola

this is a pretty good example of what I am trying to do but with better results.

 

[vk6kro]

Series resonant circuits have to have a complete low resistance path for the capacitor and inductor to exchange charge.

So, despite the name, they have to be part of a parallel resonant circuit to work properly.

For example, if you used a low impedance signal generator with 50 ohms output impedance to drive a series tuned circuit consisting of a 100 pF capacitor and a 1μH inductor, the circuit would show a very broad peak centred on 15 MHz, but 17 MHz wide, from 4 MHz to 21 MHz at the half voltage points.

Adding a 1000 pF capacitor across the signal generator's output (ie after the 50 ohms) gives a much sharper resonance.
It then centres on 16.66 MHz but has a bandwidth of only 500 KHz.
It then works as a tuned circuit and gives a voltage step-up of about 10 times.

Note that it is now a parallel tuned circuit with two capacitors in series around the loop.

These closely coupled coil experiments are really not using radio signals at all. They are just inefficient transformers.
If you align the coils so they are parallel to each other, and bring them close enough to each other, and have them both tuned to the frequency of the signal source, then you will certainly get some power transfer.

 

[jessicat]

Hallo..i have a problem..i didn't know where exactly t quote my question...so i have to design a lead compensator using bobe diagrams, for a system with transfer function G(s)=1/(s+2)(s+3) ..my problem is that the phase margin for this tf is infinite and i don't know how to find the phase tha the compensator will contribute..i have met kai solved examples where i didnt have infinite phase margin...i would be grateful if you could help me and i can explain to you the whole exercise that i have if you need further details..thank you

 

[baconman71]

That's interesting. So that means that the connections need to be in parallel for both the sending and receiving circuits?

 

[jessicat]

to be more specific ..so i have to design a lead compensator using bode diagrams, for a system with transfer function G(s)=1/(s+2)(s+3) i have the specifications for settling time and maximum overshoot but not for the velocity constant Kv..my problem is that i only have solved such problems where i had a specification for the velocity constant of the steady state and therefore i could calculate the gain in order to plot the bode diagrams as it is required in the relevant steps so as to design the lead compensator...thanks in advance

 

[jessicat]

so i supposed that the gain K has value 1 and draw the bode plots..and i found that the phase and gain margin are infinite...i have made a subject for lead , lag and lead-lag compensatin through rlocus and bode methods only in case that the compensator is in series with given transfer function..and now my professor has given me the problem that i explained to you previously ..i have solved it through rlocus methods in the time domain but i cannot solve it with the bode diagrams..i can not find what i miss ..i have even condisered whether it can not be solved without a specification for Kv because i hanen't found such a case in books or in the web...... i am really confused

 

[sophiecentaur]

http://digitalcommons.calpoly.edu/c...ontext=eesp&sei-redir=1&referer=http://schola

this is a pretty good example of what I am trying to do but with better results.

The article is quite interesting but I am really surprised that they have just loaded the resonant circuit with 50Ω across it without any apparent thought of matching or going for a high Q. i should have thought that feeding the coil with a tap near the bottom of the coil would be much more effective than hanging a low resistance across a circuit where it should be looking like a high parallel impedance. It wouldn't be difficult to sweep the resonator (using a very high resistance in series with the sig gen output and use a 'scope to find the unloaded Q. The nominal value of the C will will tell you roughly the reactances in the resonator and, hence, the equivalent shunt resistance that exists (quite a bit higher than 50Ω) and then connect the 50Ω source across a small portion of the coil turns. This will maximise the current flowing in the coil - which is, surely what you're after, if you want good coupling. It's just an autotransformer matching circuit - very old technology. Could it be that they just didn't think of this?
I am sure you would do much better with this simple mod. You could even just try a series of taps, starting at about 10% of the turns and see which gives the biggest output.
Same treatment for the receive coil, too.

 

[NascentOxygen]

Hallo..i have a problem..i didn't know where exactly t quote my question...

Hello jessicat. I suggest that you start a new thread for your questions, as a Bode plot is not related to the subject of this thread.

Once you have copied (and edited) your posts to your new thread, it would be helpful if you could delete your posts in this thread.

Welcome to PF! I hope you get helpful responses to your questions.