# Tank Circuit Problem

1. Feb 28, 2013

### mearvk

Hello all.

I'm trying to make a dead simple tank circuit and test its frequency using either my multimeter (which has the Hz measurement) or my oscilloscope.

Now I've never built one of these circuits before but it looks fairly straightforward, however I've had no luck so far in generating a measurable frequency in the LC circuitry.

My understanding is that I can use something like this:

When I try to hook this up I see 0 Hz on my multimeter.

Here is how I have my circuit: http://imgur.com/lvhgZfT

The green is the path for current to flow. The capacitor is light blue but visible. The coil/capacitor loop actually connects back via the breadboard to the place where the positive terminal connects to the coil.

2. Mar 1, 2013

The circuit will have enough parasitic resistance to decay the oscillation faster then the meter can measure the Freq - not to mention the meter has resistance that absorbs the energy as well.
What would need is a scope - you could probably get away with a PC Audio Channel scope ( http://www.zeitnitz.de/Christian/scope_en - I have not used this it was just the first in my google search) -

If you are just getting into circuits - something like this would be quite handy!

3. Mar 1, 2013

### Staff: Mentor

You'll have to feed it AC to get anything repetitive and useful for the oscilloscope. It looks like you are connecting it to DC? DC won't give you anything to see.

Have you got a widely variable AC signal generator, adjustable up to a few MHz? Otherwise, a fixed frequency square wave will do.

4. Mar 1, 2013

### mearvk

Well I hooked up a 3.82 Hz square wave to it (+5v to +0) and all I don't see anything special on the scope. It's coming out of a PIC MCU so I'm not sure exactly what the output voltage is but I'd guess 0 to 5v.

Just to be clear where should I be hooking my scope lead(s) to to verify that this thing is producing a higher/different frequency? I don't want to bother you guys if it's just that kind of issue but I've checked in a few places (within loop) and at either end of entire tank circuit with oscilloscope and nothing so far.

Thanks.

5. Mar 1, 2013

### Staff: Mentor

Can you get a square wave around 200kHz? The ringing will take place for a few microseconds or more after each square wave transition, so this will be lost from view unless you use a fast sweep. Trigger the sweep from the square-wave so the display is repetitive.

You haven't given any indication of the component values in your LC circuit, so we have no way of knowing what ballpark figure the oscillating frequency will be. What value inductors and capacitors do you have at hand?

Perhaps include a current-sensing resistor (say 10 Ohms) in series with the inductor and use the oscilloscope to examine the voltage across this resistor as you drive the circuit with a square-wave.

6. Mar 1, 2013

### mearvk

Yeah, I'll see what I can do about a faster frequency tomorrow. Should be easy enough.

Dunno exactly what the inductance is since I don't have an LCR meter but it's about 5-6 coils with a diameter equivalent to a AAA battery, made of copper wire that just barely fits into a breadboard hole (~20-22 awg); it is an air core inductor. The capacitor is in the 500 pF range. I certainly have smaller/larger caps if need be.

The FM band (88 to 108 MHz) Youtube vids all seem to work with an 5-6 coil air inductor and a pF capacitor so I'd put my circuit between 1 MHz and 50 MHz at this point, since we are working with a larger capacitor than is typical for such applications.

Thanks.

Last edited: Mar 1, 2013
7. Mar 1, 2013

### Staff: Mentor

Is your oscilloscope able to display a 10MHz oscillation? Otherwise, you could use a winding from a small power or audio transformer (salvaged from a broken cheap radio) as a large inductor.

8. Mar 1, 2013

### mearvk

I've got the OWON PDS5022T which can do 25 MHz and 100 MS/s. I'm a little unsure how to set this up best to catch the echo effect you noted, I'll play with it tomorrow and try and report back what I've witnessed.

9. Mar 1, 2013

### mearvk

I found this inductance estimator:

L = n²∗d²/(18d + 40h)

where:
L = inductance in microhenries
n = number of turns
d = mean diameter in inches
h = height of winding in inches

L = (6^2) * (0.5^2) / (0.5*18 + 40*0.375)
L = 36 * 0.25 / (9 + 15)
L = 0.375 uH

LCfreq = 1 / 6.28 * sqr_root( 0.500 nanofarad * 0.375 microhenry)
LCfreq = 1 / 6.28 * sqr_root( 0.0000000005f * 0.000000375h )

For the denominator I get: 0.000000086, check the math though.

LCfreq = 1/0.000000086= ~ 11,628,929 Hz.

Edit: re-re-tweaked some values and re-checked the capacitance which was 0.500 nanofarad, or 500 picofarad and the inductance calc was off too.

Last edited: Mar 1, 2013
10. Mar 1, 2013

### Staff: Mentor

Your 500pF capacitor is now 500nF?

11. Mar 1, 2013

### mearvk

Good catch Nascent. Changed that value back.

12. Mar 2, 2013

### mearvk

Meh. I got a 17.5 kHz signal going into the tank circuit but I'm still not seeing anything interesting on the scope. Also, my micrcontroller IDE has decided that working is too much of a big deal and so I haven't been able to get the digital oscillation speed up higher than 17.5 kHz because it is happier not working correctly.

I thought I saw some of the echo on the scope but I removed the capacitor and the signal stays about the same, so I doubt that's the LC circuit.

13. Mar 2, 2013

### davenn

well if you are feeding in 17kHz and the tank is resonant somewhere up in the MHz ...obviously it wont be resonant at 17kHz and you wont see any oscillations

Dave

14. Mar 2, 2013

### mearvk

I guess I don't really understand the correlation between input frequency and output frequency. What mathematical formula states this relationship?

My very basic understanding was that if one feeds in any AC signal that the circuit would at least create some higher version of that frequency as a product of capacitor/inductor current interchange.

15. Mar 2, 2013

### Staff: Mentor

There is none, when all you are looking for is ringing.

Your square-wave source will see those few turns of copper wire as a short circuit, and be complaining loudly. I think you should add a 100Ω series resistor (or whatever is appropriate) so that it sees a reasonable resistance.

If you are displaying on the screen a few cycles of your 17kHz square wave (it is a square wave, is it?), then you won't be seeing any of the tank circuit oscillations. You have to stretch out the sweep speed so that, if your oscilloscope screen were really huge, one cycle of the square wave would stretch across the whole room. Then, you would see some of the 10MHz ringing on the few cm of your screen. There is a one-thousand-fold difference between repetition of the square wave and the ringing (if any) of the tank circuit. To put this another way, while triggering off the square wave you need to really speed up the sweep so that the [apparently] vertical transition of the square wave actually looks more like a 45° sloping line across half of your screen.

It may take some experimenting to get a good display of the ringing. Are you using a current-sensing resistor as I suggested?

My next suggestion will be to use a hundred or so turns of wire, to make something more substantial. Then wrap a few turns of insulated wire around that to form a primary winding, and power that primary through the 100Ω resistor. The tank circuit will then not be directly connected to the square wave source.

16. Mar 2, 2013

### mearvk

The most my scope will stretch I think is down to 5ns. I had it down that low and didn't see any difference between the signal that was produced with the capacitor in place and without the capacitor in place. There was a sort of decaying sine wave at the rising and falling edges of the square wave. As I said, this stayed constant as the cap was pulled.

I have not tried putting a resistor in the LC circuit itself. Also, I am not sure what a current-sensing resistor is. I don't see a sweep feature on my scope either. I have the ability to change the time period on the horizontal scale and the voltage on the vertical scale. There may be a sweep speed function but I haven't found it yet.

I'll try tomorrow with a series resistor in the LC circuit and try and boost the frequency up higher if my microcontrollers stop dying. I got a new programmer and it seems to rejoice in lobotomizing my precious chips. :-(

I do appreciate your help Nascent.

PS I should add the smallest time period my scope runs at is 5ns. I wasn't sure if we could correlate this with a maximum frequency that it could reliably see or not.

17. Mar 2, 2013

### Staff: Mentor

That is 5ns/cm? That's plenty.
That's exactly what ringing of the LC circuit will look like, except it will change as you change L, C and/or R.
That's where the oscillations are, so if you aren't looking there, you can't hope to see any.
It's an ordinary resistor, the purpose of which is to reveal the waveform of the current in that wire.
That's what I mean by the speed of the horizontal scale.

18. Mar 3, 2013

### mearvk

Ok, latest status. 100 ohm resistor added between coil and capacitor.

It's still not working correctly but I did get a chip to output a square wave around 175 kHz.

Here are some pics: http://imgur.com/73ha149,saDv28Y,UxHXtQz#0

Yellow is the LC circuit probe, red is the square wave oscillator.

The first image is what bugs me; it shows the oscillation going into the LC circuit without the positive voltage even being connected.

The second scope image (you can see the oscillations are slightly more powerful) is with the power connected.

The final image is the two boards I have (left is LC, right is just square wave generator).

I suspect the LC circuit isn't close to being correct at this point.

Edit: I should note the relatively large capacitor was used to see if it made any difference to the oscilloscope output when compared against the very small .500 nF cap described above.

Last edited: Mar 3, 2013
19. Mar 3, 2013

### Staff: Mentor

The first image is basically what to expect. When the square-wave transitions there are oscillations in the LC circuit. Your square wave shows a sharper rise than its fall, so the tank oscillation should be more pronounced on that rise than the fall. (Remember, I said they won't be discernible until you stretch the waveform out horizontally because of their much higher frequency.)

The mains hum on your signals needs to be eliminated, and it's probably an earthing problem. What are you doing with the black earth clips on your oscilloscope probes? They should be earthed to the signal generator earth, along with part of the tank.

Try this series LC arrangement with the oscilloscope connected across the capacitor. It won't show ringing, the resistor gives too much damping, but at least it will show whether you are able to get rid of the mains hum. (If the hum can be cleaned up, try shorting out that 100 Ohm resistor and see how the waveform changes.)

I'm going to have give more thought to how the parallel arrangement can demonstrate its ringing.

#### Attached Files:

• ###### LC.png
File size:
3.3 KB
Views:
481
20. Mar 7, 2013

### mearvk

I did want to post that I have made considerable progress on this. I will post some more data this weekend if I can.

Cheers.