Amir B: More video, less comments.

I was about to write a comment about it. You need to also measure the phase between current and voltage. Actually, the precision resistor shall be characterized too. You can use two "identical" ones. 10MHz in the air has a wavelength of 33cm. So 8cm is enough to go from a valey to a peak. So everything shall be contained within a very small area so that the propagation of voltage and current waves doesn´t need to be taken in to consideration. If you measure the voltage at reference resistor over DUT, the impedance comes out of the ratio between the two waves, because Vdut/Vref=Zdut/Zref. (Same current passing through both, if wave length is greater than 1m, for instance).

Hitting near resonance . Like a kid pumping his legs on a swing. He's not putting in enough energy to push the swing that high alone, he's adding some to it on the downward stroke.

JohnTheBrewer This resistor has to be non-inductive.Ayrton-perry type wound resistors fit perfectly

That core was out of a power supply, so who knows what the properties are. :-)

So it was likely optimized for use at lower frequencies. I think your measurements are right and they show the importance of measuring a component at the frequency you will use it at. Had someone measured that coil with an LCR meter say at 100KHz and then tried to use it at say in the hf range they would likely be scratching their head .

Actually, I already started building something based on this vids idea. I have a Si5381 break out board that was donated to the channel, so I'm going to automate this measurement with an arduino nano. The board has three outputs, so it will have a few functions, including a CW beacon mode and signal generator. Doing the whole design myself, including the software.

I'll be re-visiting this in the future when I get a better signal source.

Cool, I'm looking forward to your next video about this test circuit.

Yes. I suspect that a clean sinusoidal signal source would help too. My signal was coming from a clock line on a digital chip, so it was a square wave, bent into something like a triangle wave. Very harmonic rich. I'll re-visit this method in the future whenever I finally obtain a real signal generator.

Just finished reading through your blog entry. Very extensive. And I agree that I was probably so far out of the comfort zone for that toroid I used that it probably caused the strange result. That torroid was a poor choice for the demo, it came out of an old power supply, so who knows what it's properties are. So hey, I learned something else here, thanks for the peer-review work!

Maybe. After doing a tiny bit of research I ran across the following article and learned something I'd never heard before. Each inductor has a self-resonant frequency. Below it impedance increases with frequency, above it impedance decreases with frequency. If the demo is correct the SRF of that inductor is at or below 5MHz. www.everythingrf.com/community/how-does-the-inductance-and-impedance-of-an-inductor-vary-with-frequency

The setup and math were out of a technical manual, so I trusted the source. As to Va1 and Va2 being out of phase... Wouldn't that require some kind of delay across the reference resistor? I can't imagine there would be enough delay to be measurable across a single resistor.

Kevin Loughin The current through the resistor and the unknown impedance is equal. Let's say that Zx is a capacitor, then the whole circuit becomes a low pass filter, with Va1 being the input voltage and Va2 the output voltage. This low pass filter has a phase shift between input and output, because that's a normal behavior for a low pass filter. Let's pick an input voltage of 1V and a frequency where Zx equals Rref equals 1kOhm, but as it is a capacitor, this impedance Zx is 90 degrees out of phase with the resistor. The total impedance is 1k - j1k, the absolute value 1.41k (pythagoras). The current is 1/1.41 = 0.707mA. The voltage across Rref is 0.707V, the voltage across Zx 0.707V as well, but 90degree out of phase. With your math the voltage across Rref would have been 0.293V. Hope this explains it a bit...

It's a compelling argument for sure. I'm going to re-run the experiment today with a capacitor for Zx and both inputs of the scope and look for the phase shift.

Robert why are you talking about current? you confound power with voltage. Voltage is voltage times nothing. Power is voltage times current in this case (AC) at a specific point (the same phase) But in this experiment we neither measure current nor power. We only compare two voltages regardless of the phase. Of course the waves are out of phase. But if you measure the voltage PEAK to PEAK, you "adjust" automatically the right phase. Again we are NOT measuring the product of two out of phase signals! The way Kevin do the math is 100% right. My only complain is that we need neither "precision" resistor nor "precision" oscilloscope. We can use a voltmeter too

tim46767 Hi Tim, thanks for your reply. I mentioned current only because it is the same through Rref and Zx. I don't use it to calculate power because like you wrote, power is irrelevant. But when subtracting two complex (vectors in X-Y plane) voltages like Kevin is doing, you have to take the phase between these voltages into account, otherwise you get a wrong answer. Subtracting two signals with a phase angle in between is like calculating the length of the third side of a triangle where the other two sides are the two voltages being subtracted. That is done with trigoniometry, not simply by a straight subtraction of amplitude or peak to peak values. Using a voltmeter for this measurement is only possible when the voltmeter is capable of accurately measuring a voltage at a couple of MHz. My $150 multimeter for sure can't do that. The voltmeter's AC bandwidth is not that wide.

The scope measures peak to peak, I multiply it by 0.3535 to convert to rms for calculating power in another video.

@@loughkb Why Vpp? It should be Vpp/2.