15m long RG-213 with a notch at 450 MHz

[biastee]

A ham friend gifted me with a 15m length of RG-213 coax with a strange problem. The insertion loss is within a few tenth of a dB from the specification (blue dashes) at HF and 2m (see attached graph). Above 200 MHz, the loss begins to deviate from spec. and the deepest null occurs at 450 MHz. Above 700 MHz, the loss reverts back to spec. The cable ends are connected to Telegartner BNC connectors (attached photo). I have checked the exposed shield / braid for signs of water ingress, e.g. green residue, but didn't find any. Any idea what is wrong? TIA

[edit] Return loss is better than -25 dB with the opposite end terminated with a 50 ohm load

[CJay]

Lay it out flat in one big loop instead of rolled up and see if the 'notch' moves.

[biastee]

Thanks for the suggestion.
Unfurling the cable into one large loop did not eliminate the notch, but there is an upward shift in the notch frequency (red trace in attached graph). The notch depth is reduced by ~1 dB. What does it mean?

[coppercone2]

this is the same fault I see with my LNA for a spectrum analyzer, with a notch in the same range. I still have not investigated it (its dealing with the arc of the covenant) but people were somewhat optimistic that it was some kind of issue with the internal hardline/connectors/solder. I got screwed out of the VHF-UHF transition for some reason after owning it for a while.

Since its a long cable have you tried redo the connectors?

[Joel_Dunsmore]

That's the kind of thing you can get if you have a bad spot in the cable, like someone drove across it, and it smashed down the insulating dielectric or broke the shield.  Run your hand along the cable while sweeping it and see if the notch changes at any particular point.  A break in the ground shield can act like a tuning stub put an effect short on the line (not a real short so you can't ohm it out, but a broken stretch of ground shield can act like an open which rotates to a short at some frequency).  Tough to find these things.  To quote myself  (from the book) "Cables are like dogs, either they are bad, they've been bad, or they're going to be bad.  And if they're good, they only stay good with great care"

[coppercone2]

its not hard to imagine something going wrong with the braid.

Just keep in mind that eevblog video about phase response of cables, the reflection canceling out the stimulus signal to make it notch.



The logical test would be a TDR to look for discontinuities.

[Bud]

Check your calibration by testing a device (or a shorter cable)  with a known frequency response in 300-600MHz range, maybe your Thru cal standard is defective.

[TheSteve]

Time to break out TDR perhaps - if it can't find the fault then ditch the BNC connectors. Well I'd ditch the BNC connectors anyway, they are not good at all for physical loads on the connector and 213 is heavy and stiff.

[biastee]

Since its a long cable have you tried redo the connectors?

Surprisingly, the Return Loss (RL) is good even at the afflicted frequency ~450 MHz (black trace in attached graph).

I didn't redo the connectors but I did disassemble one connector to check the braid and the soldered centre pin - but found nothing suspicious. 

Would TDR be able to reveal the fault location if the RL (S11) is good?

Run your hand along the cable while sweeping it and see if the notch changes at any particular point. ..

Thanks for the useful tip. I plan to do this in the coming days.

[edit ] "Cables are like dogs". Ah! "3.13.1 Cable Flexure", pg197!

Check your calibration by testing a device (or a shorter cable)  with a known frequency response in 300-600MHz range, maybe your Thru cal standard is defective.

I measured a LMR-400 of similar length and the result was unremarkable (attached graph). So, I don't suspect the Thru standard is faulty.

Thanks to all for the tips!

[CJay]

Thanks for the suggestion.
Unfurling the cable into one large loop did not eliminate the notch, but there is an upward shift in the notch frequency (red trace in attached graph). The notch depth is reduced by ~1 dB. What does it mean?

Depends, can you compare to a known good cable of the same length?

It could be a faulty piece of cable, badly made connections, a problem with your VNA, the calibration, the connectors or the connecting cables.

Did you run the calibration process, what happens if you run it again but narrow down the span of the VNA to 250-650MHz?

[A.Z.]

Did you try REVERSING the coax connection and checking if the "dip" remains at the same frequency ? You may either terminate the other end of the cable to a good dummy load (going up to that freq) or just short the other end

[biastee]

Depends, can you compare to a known good cable of the same length?

It could be a faulty piece of cable, badly made connections, a problem with your VNA, the calibration, the connectors or the connecting cables.

Yes, I did measure a 15m length of LMR-400 and the result was unremarkable - please see the graph in my previous post. I have also measured a 2 dB attenuator (Kay 1/839) and there was no notch (graph attached). These two results convinced me that the VNA (nanoVna V2+) and its calibration are satisfactory.

Did you try REVERSING the coax connection and checking if the "dip" remains at the same frequency ? You may either terminate the other end of the cable to a good dummy load (going up to that freq) or just short the other end

The cable was measured in the transmission mode;i.e. connected to both VNA ports. Reversing the cable did not change the notch frequency.

Thanks to all for the responses!

[gf]

Would TDR be able to reveal the fault location if the RL (S11) is good?

What hinders you to turn on the time domain display and take a look?

[PE1MHY]

I have seen this kind of behavior with a bad production RG213 where the dielectric material was contaminated. But to really notice this the cable had to be much longer.
In your situation I would first look at the measurement and the calibration. For me your RL result is just to good and looks like the calibration procedure correction-factor combined with the cable.
When you are sure about the calibration and your RL measurement (test it with good cable) and the result still looks like before I would assume the cable is mechanically distorted or damaged.
   

[biastee]

What hinders you to turn on the time domain display and take a look?

I did investigate using the TDR low-pass step function, but didn't mention it in my original post because I didn't know what to make of the results. The attached results are for open and short conditions at the opposite end of the cable (attached graphs). Sorry for submitting screenshots instead of captured data - I don't know how to use the supplied PC software, vna qt's / vna view's TDR function.  When the cable's opposite end is terminated in 50 ohm, the TDR result is simply a straight line.

When you are sure about the calibration and your RL measurement (test it with good cable)

Yes, I have tested a good cable - a 15m length of LMR400 clone. Please see the graph attached to reply #8 in this thread.

[cdev]

TDR should be able to tell you within a very exact amount, where whatever is causing this is, in terms of length. That is "the best" way to do this. Since it tests the core function of the transmission line.

Are you sure that there is not any water in the cable from being in the rain? Water will cause problems like the one you describe.

I would - if you get an indication of a point where there is a discontinuity- would also measure the cable with a caliper or digital micrometer if you find that the cable is not circular anywhere that probably represents a change large enough to cause the loss.  Foam cable can deform so much sometimes that it has much more loss.

Ask any "Cable guy"

[gf]

I did investigate using the TDR low-pass step function, but didn't mention it in my original post because I didn't know what to make of the results. The attached results are for open and short conditions at the opposite end of the cable (attached graphs). Sorry for submitting screenshots instead of captured data - I don't know how to use the supplied PC software, vna qt's / vna view's TDR function. When the cable's opposite end is terminated in 50 ohm, the TDR result is simply a straight line.

A horizontal line at zero is fine, for a 50 Ohm terminated cable.
But open and short look weird. Does the good cable look the same? If it does, then I would not rule out that something might be wrong with the measurement setup or calibration.

Did you consider that start frequency must be 50kHz for lowpass mode (i.e. the lowest possible frequency -- ideally it should be zero)?
Did you re-calibrate the NanoVNA after changing start/stop/#points? (directly on the device, since calibrating the PC software does not affect the device)

[biastee]

Are you sure that there is not any water in the cable from being in the rain? Water will cause problems like the one you describe.

I would - if you get an indication of a point where there is a discontinuity- would also measure the cable with a caliper or digital micrometer if you find that the cable is not circular anywhere that probably represents a change large enough to cause the loss.  Foam cable can deform so much sometimes that it has much more loss.

How to determine whether there is water ingress? What I did was to open up one BNC connector and then inspected the exposed copper braid for green residue, but didn't find any.

The cable is RG-213 which has an insulation of solid PE, NOT foam.

Did you consider that start frequency must be 50kHz for lowpass mode (i.e. the lowest possible frequency -- ideally it should be zero)?
Did you re-calibrate the NanoVNA after changing start/stop/#points? (directly on the device, since calibrating the PC software does not affect the device)

Yes, I did set the start frequency to 50 kHz because that requirement was clearly stated in the nanoVna V2+'s instruction manual (pg. 19). However, I didn't re-cal after changing the start & stop frequencies. I thought it was unnecessary because of two reasons: 1. the manual's TDR section didn't mention calibration and 2. the classic TDR (square wave gen & scope) didn't require OSL calibration. Anyway, the results after OSL cal are attached below. Can you see anything that would explain the 450 MHz notch?


Thanks to all those who responded!

[gf]

All kind of VNA measurements require proper calibration. Quote: "Calibration must be performed whenever the frequency range to be measured is changed." If you do TDR with a VNA, then S11 and/or S21 are still measured in the frequency domain, and the corresponding time domain data are calculated via IFFT then. The calibration data are stored per frequency point (for those frequencies which are in effect when the calibration is done). So when the frequency points change, the calibration data for the new frequencies can only be interpolated from the old calibration, or even just extrapolated, so that accuracy suffers. So better always re-calibrate when you change start/stop/center/span/#points, or when you change the set (which includes all involved cables and/or adapters up to the calibration planes). You can also restore a saved calibration, which restores start/stop/#points in together with the fitting calibration data. Still it is up to you to ensure that the test set is the same as when the saved calibration was established. Since (at least my) NanaoVNA does drift, I also do not rely on stored calibrations for a longer duration.

EDIT: I see you added images. Much better now. It seems, though, there is an impedance mismatch already at the source, which would imply that the cable impedance were not 50 Ohm. If you switch the trace type to IMPEDANCE RESISTANCE then you can see the impedance profile over time (distance).

BTW: The start is outside the left border of the screen. Try to shift the traces a small amount to the right by applying say -20 ns Electrical Delay. I guess the horizontal part of the trace covering the cable length will go even a little bit more negative then.

How does it look with the "good" cable, for comparison?

[Bud]

I am thinking the RG213 may be leaky and what you see is a result of coupling the leaked fields to the surrounding environment. LMR400 is a double-shielded cable so leakage is much smaller. To test this theory put a few snap-on ferrite chokes on the RG213, at the ends and along the length, and see if that affected the resonance amplitude or frequency.

[biastee]

EDIT: I see you added images. Much better now. It seems, though, there is an impedance mismatch already at the source, which would imply that the cable impedance were not 50 Ohm. If you switch the trace type to IMPEDANCE RESISTANCE then you can see the impedance profile over time (distance).

BTW: The start is outside the left border of the screen. Try to shift the traces a small amount to the right by applying say -20 ns Electrical Delay. I guess the horizontal part of the trace covering the cable length will go even a little bit more negative then.

How does it look with the "good" cable, for comparison?

When the format is switched over to "Resistance", the cable portion was ~ 40 ohm. Attached photos of the bad RG-213 cable and a good LMR-400 clone - both with "OPEN" at their opposite ends. A -20 nS delay has been added. The LMR-400 appeared shorter than actual because the propagation velocity was at default 67%; i.e. not changed to 85%.

I am thinking the RG213 may be leaky and what you see is a result of coupling the leaked fields to the surrounding environment. LMR400 is a double-shielded cable so leakage is much smaller. To test this theory put a few snap-on ferrite chokes on the RG213, at the ends and along the length, and see if that affected the resonance amplitude or frequency.

I don't have snap-on ferrite chokes that are large enough to fit the RG-213's 10mm OD. Can I use TDK T19-31-8 (ui = 290) toroid rings which is large enough to slip over the BNC connector instead?

Thanks to all for the responses!

[Bud]

Use whatever magnetics you have , the point is to test the leakage idea, whether using magnetics affects the resonance.

[coppercone2]

you know, how about testing that wire with a megger? If its contaminated dielectric maybe the leakage current will be higher.

Measure shield to center pin resistance, preferably first sweep it at a lower voltage (100V?) with a electrometer if it does not reveal anything, plug it into a megger.

[biastee]

To test this theory put a few snap-on ferrite chokes on the RG213, at the ends and along the length, and see if that affected the resonance amplitude or frequency.

No joy with the ferrite rings (see attached photo).

you know, how about testing that wire with a megger? If its contaminated dielectric maybe the leakage current will be higher.

Measure shield to center pin resistance, preferably first sweep it at a lower voltage (100V?) with a electrometer if it does not reveal anything, plug it into a megger.

:-) I really didn't expect to hear "megger" on an RF forum! Ok, I am throwing in the towel at this point! I don't own one and I don't intend to splurge on one for the following reasons: 1. this test is probably not specific because a crushed cable would also exhibit a lower breakdown voltage, and 2. a positive diagnosis will not enable the problem to be rectified.   

Thanks to all for the responses!

[coppercone2]

I mean, you can just put any voltage across it and measure current. does not need to be accurate just compare to another cable.