How to measure transmission line?

[K9OM]

Hi Guys,

Is there a way to measure the Impedance and loss in a transmission line if you don't know the make/model/velocity factor of the cable?

In my case, I need to measure the impedance and line loss of a 50 year old transmission line that is aprx. 500' long, buried, and feeds an AM broadcast tower.

The only things I know about the old transmission line are:
-aprx. 500' long.
-aprx. 5/8" in diameter.
-has a heavy rubber outside jacket.
-the braid is copper and inner conductor is stranded copper.

I have no idea what the impedance or velocity factor is.

All thoughts appreciated!

Thanks!

[w0qe]

This is an easy question BUT before I answer do you have access to both ends of the piece of coax? In other words can the far end be disconnected and do you have access to it?

If you do then you can measure attenuation vs frequency, characteristic (surge) impedance of the coax, but you can't determine both the physical length and the velocity factor. A 500' roll of .66VF cable appears to be electrically 750' long which is exactly the same as 615' of .82VF cable.

73, Larry W0QE

[K9OM]

Hi Larry,

Yes- I do have access to both ends of the transmission line. Do you recommend that I test the line: Open or Shorted?

What I really need to know it the actual line attenuation at 1.370 Mhz, and if possible, the line impedance at 1.370 Mhz.

Thanks,
thingy- K9OM

[Bob]

You can read the cable loss at 1.37MHz directly from the AIM data shown on the right side of the graph. This is valid when the cable is either open or shorted. It assumes the magnitude of the reflection coefficient at the far end is 1.0. Do a scan and then position the cursor to 1.37.

The impedance of the cable, Zo, can be calculated by measuring Zin (Z at the input end of the cable) with a short and with an open at the far end:

Zo = SQRT(Zin_short * Zin_open)

The Z's are all complex numbers. This assumes the cable is uniform and there are no stubs or filters involved.

73/ Bob

[w0qe]

Ok Thingy (I'm sure you have another name ),

Let's start with what the coax might be. Look at www.timesmicrowave.com/cgi-bin/calculate.pl for a good example of what coax is available and what the loss and power handling capabilities are. My guess for your coax might be RG-217 or LMR-600-UF.

RG-217 (50 ohm) at 1.37MHz & 500' matched loss = 0.8dB and avg. power = 14.6kW
LMR-600-UF (50 ohm) at 1.37MHz & 500' matched loss = 0.5dB and average power = 21.5kW

I assume you are wishing to reuse the coax if the loss is acceptable and you have already determined the coax to be physically in good condition.

Let's now define the problem:

1.) Measure the characteristic impedance of an unknown piece of coax.
2.) Measure the loss of the coax which is expected to be between 0.5dB and 3dB.
3.) Make the measurements with a 1 port network analyzer (AIM4170 or PowerAIM).

To start, do a custom calibration on the AIM4170 or PowerAIM (just AIM from now on) at the end of a reasonable (<20 ft.) piece of coax that you will connect to the big coax later. I would do a custom calibration with a starting frequency of 1MHz, a stopping frequency of 10MHz, and a step size of 0.1MHz or a couple hundred steps. If you will not be using connectors to connect to the large coax you can easily have a couple inch long pigtails on the short coax opened, shorted together, and a 49.9 ohm 1% resistor (short leads) for calibration. Or you can use some other value resistor in the range of 40 to about 150 ohms (as long as you know the value to 1%). Set the AIM to make SWR and return loss for a 50 ohm system (to start).

Measure the coax characteristic impedance:

Method #1: Connect the short coax to the big coax. Open the far end of the big coax and set the AIM to pretty much any frequency. Measure the impedance and call this value #1. Now short the coax and at the same frequency measure the impedance and call the value #2. Both these numbers are complex. Multiply value #1 times value #2 and then take the square root (Yeah, I know you can't easily tak the square root of a complex number) but by converting to polar coordinates and taking the square root of the magnitude and dividing the phase angle by 2 achieves the same thing. If you look at the number you are trying to take the square root of it should be nearly real since most coax has a characteristic impedance that is something like 50 +j0.5. You may wish to do this calculation at 10 MHz (characteristic impedance of coax varies very little with frequency and the numbers may be a little more accurate since the coax will have more loss at 10MHz.

Method #2:
Terminate the far end of the coax with a guess value of the characteristic impedance. My bets are for 50 ohms. Sweep the AIM from 1 to 10MHz and the impedance measured should be very close to 50 +j0 for all frequencies. If you don't get 50 + j0 you can then try another value such as 75 ohms and redo the test. Or you can plot all the values anf the geometric center of the impedances plotted on a Smith Chart will be the answer.

Now to measure the loss in the coax:

For this test the AIM needs to be set to the impedance you determined the coax to be and at a frequency of 1.37MHz. Open or short the far end of the big coax and measure the return loss (SWR and return loss are 2 ways to measure the same thing which is reflection coefficient). Open or shorted coax will reflect 100% of the RF so the attenuation will be 1/2 of the measured value. The AIM is a really poor device to measure coax loss if the loss is only a fraction of a tenth of a dB or less since the returned SWR is over 100:1. In a case such as that a 2 port network analyzer is needed. However here you expect a minimum loss of 0.5dB (meaning a 1.0dB return loss) which is the same as 17.39:1 SWR which is quite achievable with the AIM. Alternateively since the loss in coax (at HF) increases as the square root of the frequency (refer to the loss calculator above) you can measure the loss at 4 times 1.37MHz and then divide the attenuation in dB in half which is why I suggested to a cal. up to 10 MHz.

Hopefully this helps. There is plenty of theory to support the mentioned tests and I have done these exact tests many times. Good luck and if anyone knows a better way to do the tests please speak up.

73, Larry W0QE

[w0qe]

Darn,

I started my reply before Bob replied but got sidetracked by dinner so much of what I said is what Bob said. Nevertheless with 2 people saying something perhaps it is easier to understand.

73, Larry W0QE

[K9OM]

Bob & Larry,

Thanks a bunch for the detailed explanations. I'll take that information to the tower site this week and "go at it like Abe Lincoln". That's kind of a popular saying round these parts as WLLM-AM is located in Lincoln, Illinois. (the first city named after Abe)

I sure don't know how the name "thingy" got in my post... but it made for some fun)

Best 73,
thingy- K9OM (aka "thingy")

[w0qe]

AKA Thingy,

Good luck with the measurements and please let us know how it went and feel free to share the data if you wish.

In re-reading my reply, I made the comment that the AIM was a poor choice to measure very small losses in the cable. This comment is in no way directed to the AIM directly but to all single port network analyzers which nust use an impedance measurement (calibrated to the Zo of the coax) to infer insertion loss. A 2 port calibrated network analyzer, of lesser quality, can much more easily measure amplitude differences of a few hundredths of a dB compared to a single port analyzer trying to measure super high SWRs. E.g. 0.01dB cable attenuation equates to an SWR of 868:1 and a 0.02 dB cable attenuation equates to an SWR of 434:1. Trying to make measurements like this is at best foolish. Often you might be tempted to measure the loss of a short piece of coax and extrapolate up to large lengths which needs a 2 port analyzer for any real accuracy.

In your case, since you are measuring the full cable length, you should have no problem resolving the difference between 0.5 and 0.6dB. I just made a couple of quick measurements on the AIM4170 with 2 resistors (866 and 732 ohms) which represent example impedances that could be produced by SWRs or 17.4:1 and 14.5:1 which correspond to cable attenuations of 0.5dB and 0.6dB. The AIM easily resolved these values with good accuracy.

All measurement devices have strengths and weaknesses. Understanding how to measure the desired property most accurately given the test equipment available is an art and forces you to understand the underlying theory but that is half the fun.

73, Larry, W0QE