Learning LC Circuits The Hard Way: Rolled Up Aluminum Foil And Wax Paper Scraps
This will not be an easy post to write. There are far too many RF and electrical concepts cemented together with way too little foundation of knowledge on my part. It is a continuation in some ways to this post. And a bit to this post. And it fueled by some challenges posed by my friend and critic 😉 AC1JR. His blog is here.
So, the guts of my band pass filters seem to just be coils and capacitors. This latest adventure started with thinking “Boy, I could make this myself!” It is also a real exercise in understanding the importance of inductors and capacitors in an AC circuit. Not in the oh, this is just another tidbit of knowledge to take in. But more of an internalization that LC circuits are the magic wand of RF. Maybe the bedrock of RF. Every road I take is leading me back to these resonant little buggers: baluns, antennas, tuners, and now the filters…
Ready for this journey? A warning, it is a lot more like a meandering brook than a linear train track… So, I watched this video, and was quite amazed by the idea of homemade components. It led me to trying this:
No surprise that the coil had some inductance (0.14 µH). But the fact that the washers actually had 3pF of capacitance surprised me. And if you use a LC resonant frequency calculator for series inductor and capacitor band pass filters like this one, you can find that the F0 (i.e., fundamental frequency) of the circuit is 246 MHz. Or go old school: F0 = 1/[2π√(LC)] with L in henries, C in farads, and frequency in hertz. As you can see, I did build the circuit on a breadboard, and then sent a sinusoidal signal through it and watched it on the oscilloscope as I increased the frequency. I saw some attenuation of the signal voltage occur, but nowhere near 246 MHz since my signal generator only goes to 65MHz.
And much like a cat finding a new ball of string, I was distracted away from this particular circuit by a correspondence with AC1JR. Apparently, you can make a capacitor out of aluminum foil and paper. Oh my goodness!
That first photo is a sheet of wax paper folded over a strip of aluminum foil. And another strip of aluminum foil on top of the wax paper. The aluminum strip is left long at the left hand side for illustration purposes only; the aluminum should be entirely contained within the wax paper so that the two aluminum strips remain separated from each other. Then the entire paper & foil sandwich is rolled up. When you get toward the end, simply use a piece of electrical tape to adhere a bare piece of wire to each sheet of aluminum and arrange it so those wires stick out of the end of the roll, making sure they do not touch. I then went ahead and wrapped the entire thing in electrical tape. No joke, a capacitor!
In fact, it has 46 nF of capacitance!
And, remember when I built my not-quite-functioning receiver? Well, I repurposed the coil for this project. It registered as 1.33 µH of inductance.
Therefore, the fundamental frequency of a series LC circuit (i.e., a band pass filter that will pass frequencies around the fundamental and block other signals away from the fundamental) made of these two DIY components would be 644 KHz.
I made a series circuit out of my DIY components. I do not know if it affected the results, but I fed the signal in through the capacitor first. You may be able to see that I soldered smaller gauge wires on to the ends of the coil so that it would fit in to the breadboard.
I also soldered wires to the SO239 connectors so that I could use them in breadboard as well.
And then I gathered my nanoVNA, SMA-to-PL259 test coax, and open-short-load calibration set (purchased here). The nanoVNA was used with the open source (thanks AC1JR!) software NanoVNA-Saver.
My first line of business was to calibrate the NanoVNA with the test leads attached. I used the NanoVNA-Saver calibration wizard which walked me through the process (shown below).
After the calibration was complete, I hooked up the CH0 test lead to the input SO239 connector of the circuit and the CH1 test lead to output SO239 connector. I did not know what the calculated fundamental frequency was when I first ran these tests, so I just started with 1MHz and went to 1GHz initially. The peak/dip marked with arrows caught my attention, so I focused in on that.
My next sweep looked at 1MHz to 200MHz. The peak around 170MHz was identifiable again. You can see the triangular markers around the peak/dip of interest that the software added in.
The NanoVNA-Saver has an “Analysis” feature. In it I chose to look at the characteristics of this region as a band pass filter. The results are shown below, with the fundamental frequency in this region calculating as 166 MHz with a 6dB span from 147 to 187MHz. This of course is in contrast to the mathematically calculated fundamental frequency of 644KHz. I am at the limits of my knowledge in understanding why there is resonance at this particular frequency. If you look at the lower frequencies on the plot above, you will see that the NanoVNA has a “shotgun” pattern of results, so looking at that area of the curve (i.e., frequencies <1MHz) were uninstructive, it was just a smattering of scattered seemingly random results.
And finally, here are some closer views of the return loss and VSWR of my “band pass filter” from 130 to 200 MHz.
During all of this testing, however, I noticed my scanner kept pausing on “VHF High Wireless Microphones”. I had never seen this before. In fact, it freaked me out. I specifically talked to see if I would hear myself in the scanner! Did someone put a wireless microphone in my qth?! All I could here coming from the scanner was nothing but dead air. It was creepy enough that I sent a picture to AA1F.
I put away my experiments for the evening and started the pre-bedtime routine. And then it dawned on me… I pulled up the picture of the scanner and looked back on the VSWR graph…171.845MHz. I had been transmitting! Uh oh, out of ham band despite what was surely a tiny amount of RF that was going no farther than my workbench, but still…
So how could I prove this? Well, the next night I moved the experiment into the faraday room in my basement. Ok, maybe an exaggeration, but I wanted to try one more thing before I put this away for good and I did not want my signal to transmit anywhere if it indeed was transmitting so I figured being surrounded by concrete blocks would help. I set up the NanoVNA to the LC circuit identical to the way that I did above. I also had my SDRPlay RSP2 with spectrum analyzer software running. I was analyzing frequencies between 155 and 205MHz. I checked how long it took the NanoVNA to sweep between these frequencies (using 50 segments as the settings); it took 1 minute and 16 seconds.
So my first test was to run the spectrum analyzer for 1 minute and 16 seconds to capture ambient frequencies. I used the “peak” function which would show all of the peaks detected in that time period. Here it is:
Next I connected the two test leads with a barrel connector and ran the sweep through the same frequency range on both the NanoVNA and RSP2 spectrum analyzer. This is of course to control for the NanoVNA itself transmitting. Here are those results, with the NanoVNA VSWR graph now superimposed in the top left corner. The spectrum analyzer results look fairly similar to the ambient environment.
Finally, I plugged my homebrew kinda band pass filter in to the NanoVNA in the same way I had done earlier. I ran a single sweep from 155 to 205 MHz. Here are those results:
A much different picture! In fact, as the plot for the VSWR was formed, you could watch the intervals of the spectrum analyzer pick up new peaks simultaneously. If you really strain your eyes, you can see the inverted “U”-shape of the signal peaks, much like the reciprocal of the VSWR plot. So, my band pass filter, well, was no such thing. It is as best as I can tell, actually a VHF transmitter. It is now in the trash never to be used again. Hard to believe though that my first homebrew transmitter was rolled up wax paper and aluminum foil and a coil of wire…
Best,
KM1NDY