Great vid: Why doesn't a 75 Ohm cable measure 75 Ohms?


The most complicated way to describe is a TDR, Time Domain Reflection, impedance. Yes, it does work this way and their are devices that are made to do just this test called, of course, a TDR. This does cover a lot of RF trickery that is nice to know but for just measuring impedance it is the most complicated way to do it.

What if we don’t have all the fancy stuff? What’s easier for determining a cable impedance?

-) You could use the equation to calculate the RF impedance. You need some dimensions and a few factors on dielectric stuff to do that but no electrionics.
-) You can use the Belden catalog on a Belden coaxial part number and use the simple equation 101670 / capacitance * Vp). The cable specs will give both those numbers. Example 101670 / 17.3 * 78 = 75.3 ohms.
-) You can take a known 75-ohm cable and measure the conductor’s diameter and the insulation diamater under the shield and get the RATIO of the two. If the DIELECTRIC is the same material, the ratio will remain the same across any cable of the same impedance.An RG59 and an RG6 with a foamed dielectric of the same Vp will be the same RATIO.
-) We can also use the cable data and use the equation; SQRT (L/C). The L and C have to be in henries and farads, yes, with all the zeros in there. 17.3 pF needs to be 17.3E-12 for example. Often the inductance isn’t available as the 101670/ C*Vp is so easy to use and more accurate as L is a harder number to measure right.

Harder as you need some fancy stuff;
-) You could use the technique in the video and change the termination resistor until the reflection is at a minimum. This is the characteristic cable impedance when that happens. This is how RETURN LOSS, RL, is calculated. You put a fixed resistor at the cable’s end that’s supposed to be the target impedance, and measure the return loss across FREQUENCY. Ideally there should be none at any frequency. Cable isn’t perfect, so the “impedance” will vary as will the reflection in dB.
-) You can measure SRL, STRUCTURAL RETURN LOSS, by altering the termination impedance until at any given frequency the return loss is MINIMIZED by the cable’s structure at that specific frequency. We cheat and alter the load to match the cable, not the other way around.
-) You can us the open-short method. This uses the measured data and an equation to calculate impedance. Zo = SQRT (open*short). It is also useful to get the reactive angle of the impedance as this tells use how good the RL will be. We want a zero degree phase angle (all resistive). We won’t get that, though. It is a quick way to estimate RL.

This stuff also works for twisted pairs, like Ethernet cable.For example 100-ohm twisted pair is two 50 Ohm coaxial cable cores in parallel. The distance between the two center wires DOUBLES, halving the capacitance and this raises the impedance. The 101670 / (C * Vp) tells us that. If the Vp is the same and C goes down, Impedance goes up.

All the methods are inter-related. They just use different knows to get the unknown. I’ve used them all depending on what you have at the time, and what you want to know. Just impedance and RL? Impedance and phase? Actual cable impedance (SRL) at a frequency?

Best this is all much more controlled at RF as Vp, L and C is constant across RF and technically phase is near zero. The cable looks like a resistor. All we need to do, easy to say, is make the cable the same reactive value at all frequencies so we just have attenuation to deal with. Ya, sure, that’s all!


Without years of practice and real world application of these equations and terms, it’s really hard to get a grip of what is actually going on, with or without a scienticfic background.

I found this disconnect when diving head in into the hi-fi hobby. After a certain point, there is no clear cut understanding from the public domain as to what scientific term or a combinations of them contribute to what we hear and not hear. All there is that I know works, is to throw shiit on the wall and see what sticks. Shiit being my $$$.

This is too inefficent, even with the helps of fourms like PS Audio, the signal to noise ratio is simply too low.

Would love this industry has more videos like the one Iinked, Paul’s videos and @rowr30s presenations for those that want to understand more.

1 Like