I have a lot of 'donated' coax that has been given to me with unknown provenance. I'm wondering what things I can do that can determine whether I keep it or send to the tip? Some of the things I have thought about doing are:

• Physical inspection - look for oxidisation of shielding braid, damage to plastic sheath
• Continuity testing, short circuits between inner and outer.
• plug into dummy load and do an SWR sweep
• Measure cable loss and compare against specifications

The big one is cable loss - Any thoughts on how I can measure that cheaply? I don't have any calibrated test equipment for that?

Any other tests I should think about?

You're on the right track with the first items in your list. Assuming that you have a wattmeter and a dummy load that both match the impedance of the coax, it's a simple matter to measure the loss.

Measure the power with the wattmeter at the source, and then measure it again at the load. The difference between the two wattmeter readings is your loss.

$$N_{dB} = 10\log_{10}\left(\frac{P_2}{P_1}\right)$$

Note that this technique depends on the calibration of the wattmeter, and does not itself detect if the coax has damage leading to reflection rather than dissipation (you would want to verify the SWR is as good as without the coax, before assuming the watt figures are good).

• Note that this technique depends on the calibration of the wattmeter, and does not itself detect if the coax has damage leading to reflection rather than dissipation (you would want to verify the SWR is as good as without the coax, before assuming the watt figures are good). – Kevin Reid AG6YO Nov 25 '18 at 22:11
• @KevinReidAG6YO Excellent! I copied this to my answer. – Mike Waters Nov 25 '18 at 22:20

A cursory physical inspection and an SWR sweep are usually sufficient.

Blatant physical damage, like internal shorts or breaks in continuity, will be found by the SWR sweep. So in your physical inspection you're looking for things which might not impact the SWR, like outer insulation that's cut or damaged which may eventually lead to water ingress.

You could measure loss, but I can't think of any likely mechanisms of loss that wouldn't also lead to anomalies on the SWR sweep, so I wouldn't worry about it too much.

In my experience, the problems with salvaged coax are most usually the connectors. Hams are notorious for buying cheap connectors and then installing them incorrectly. I'll wiggle the connections while the SWR sweep is running: any mechanical issues will show up as extreme spikes in the SWR. If the connectors are at all questionable I'll install new ones, and this also affords an opportunity to inspect the conductors for oxidization, an indicator of water ingress.

• An easy way of detecting loss if one can measure impedance, e.g. with an antenna analyzer, is to measure the impedance of a shorted or open stub and compare it to the results predicted by TLDetails available from AC6LA.com. If the results are appreciably different from those predicted by TLDetails, that coax would probably make better radials than transmission lines. – Cecil - W5DXP Dec 1 '18 at 18:19

In addition to the above tests, and if:

• the coax is of some reputable brand and has type markings on the jacket,
• the cables are of sufficient length,

you might want to check the capacitance and compare to the capacitance in the datasheet. This is usually marked as F per meter/100 meter/feet/... Calculate the expected capacitance for the length you have, and compare to the measured capacitance between jacket and the center. If there is a large amount of moisture ingress (for example after prolonged outdoor use), this could affect the capacitance.

This is also a test that can help you get a rough estimate of the length of longer lengths of (installed) cable. I have used this at work to check an expected 285m run of coax cable installed on the winch of a ship. Somewhere in the signal path we had a problem, but access to connections was difficult. By measuring the capacitance (in addition to a pulse reflection test), we found that the cable on the winch was the culprit.

I'm trying to solve a problem right now that bears on your situation.

Assuming we know what frequencies you're interested in using the cable for, and that you have a signal generator, receiver (CW is best) and a terminator of the appropriate impedance:

Set up the signal generator on a frequency roughly in the middle of your range, output 0dBm or higher. Terminate the far end of the cable. Tune the reciever to the same frequency. Run the reciever along the coax, defects in the shield (or really bad shields) will show up as a loud signal.

A "perfect" cable would have no detectable leakage other than that dictated by its transfer impedance: https://www.araconfiber.com/what-is-transfer-impedance/

It's not a quantative test, but it's informative.

I have some Pasternak RG-58 "100% shield" cables here that are VERY leaky. I have some Andrew RSJ1-50 cables with leakage that's almost undetectable except where the connectors attach to the cable (factory mounted, not my work) Very noticeable difference between the two.