Someone at my amateur radio club asked me if I could measure the impedance of some coax. which he purchased from a retailer of amateur radio equipment. It was sold as 50 Ohm, but looks like cheap 75 Ohm TV coax to us both. Personally I would chuck it in the bin, but I said I will measure it for him. I have a few metres of the stuff.

What is the best way to measure the impedance of this?

I know its possible to do things with the time-domain option on a VNA, but I'm not sure what frequency range to use. Two issues make me think that going from DC to light is not ideal

1) I know the impedance of coax rises above the high frequency value, so would collecting data at less than 10 MHz result in me giving him the wrong nominal value.

2) I also know that this guy will not use this above 432 MHz.

So is there any point sweeping it over a frequency range below 10 MHz, or above 432 MHz? The fact the stuff looks pretty cheap and horrible, makes me think it is unwise to test it at a high frequency. If it is 1 Ohm or 1000 Ohms at 10 GHz, it is going to make no difference to him.

But I only have a sample that is a few metres long.

I have at my disposal.

* 8720D with time domain option. But the lower frequency limit of the 8720D (50 MHz) is higher than the frequency of interest to him.

* 8753ES. The frequency range of this (300 kHz to 3 GHz) more closely matches the frequency range he would use this stuff at. But the 8753ES lacks the time-domain option. But of course I could do the inverse Fourier Transform outside the VNA - the software for the low-cost VWNA will do the IFT from a Touchstone file.

Dave

What is the best way to measure the impedance of this?

I know its possible to do things with the time-domain option on a VNA, but I'm not sure what frequency range to use. Two issues make me think that going from DC to light is not ideal

1) I know the impedance of coax rises above the high frequency value, so would collecting data at less than 10 MHz result in me giving him the wrong nominal value.

2) I also know that this guy will not use this above 432 MHz.

So is there any point sweeping it over a frequency range below 10 MHz, or above 432 MHz? The fact the stuff looks pretty cheap and horrible, makes me think it is unwise to test it at a high frequency. If it is 1 Ohm or 1000 Ohms at 10 GHz, it is going to make no difference to him.

But I only have a sample that is a few metres long.

I have at my disposal.

* 8720D with time domain option. But the lower frequency limit of the 8720D (50 MHz) is higher than the frequency of interest to him.

* 8753ES. The frequency range of this (300 kHz to 3 GHz) more closely matches the frequency range he would use this stuff at. But the 8753ES lacks the time-domain option. But of course I could do the inverse Fourier Transform outside the VNA - the software for the low-cost VWNA will do the IFT from a Touchstone file.

Dave

The impedance of a piece of coax changes along the coax, and at each position in the coax, changes with frequency; thus, assigning a single value implies some averaging of results.

The time domain transform will show you the variation average impedance (averaged across the frequency used to generate the time domain) along the lenght of the cable.

The S11 return loss, converted to impedance through well known formulas, will show you the average impedance (averaged along a portion of the length of the cable, that portion being dependent on the S21 loss of the cable) provided that the cable is terminated in a impedance load matched to the cable. Alternatively, you can adjust the source and load impedance of your measurement system (modern analyzers have port impedance transformation) to minimize the S11 response, and reprort that source and load impedance as the impedance of the cable (It is the impedance which best matches the cable for minimum return loss).