Measurement of cable loss in with high VSWR

[jkkpohl]

high vswr cable power loss

Is any one familiar with how to measure cable loss in a high VSWR environment? The losses will go up due to increased current standing waves but my method of using a directional coupler at the input and output end of the cable does not reflect that.

 

[rigdoctor99]

vipec example

What you need is an SWR meter, this will show forward and reflected power........... and if your antenna has a high VSWR you should stop using it and tune it for a lower SWR...... google search will give you lots of ideas........... You dont say what frequency you are working. Let me know, I may already have the data you need............

 

[Gakusei]

loss due to vswr ripple

This is quite a common problem. Because your cable has a high return loss (I,m sorry I can't think in VSWR), your input directional coupler is picking up a mixture of the forward and reflected voltage. As these can be in or out of phase depending on the reflections recieved, you results will be meaningless.

The solution is to use a directional coupler with higher directivity, better than 40dB if you can find one.

 

[jkkpohl]

how to measure of high vswr

I have characterized these directional couplers to have very good directivity on the network analyzer. With either and short or open the coupling factor does not change more than a few tenths of a dB. I have actually "load pulled" the output load and still achieved no worse than a few dB error for a large locus of output impedances.

It interesting to note that in a high VSWR situation that the directional coupler may have trouble separating incident from reflected voltages. I am consider making direct voltage or perhaps peaking power measurements with a high impedance probe instead.

 

[sergio mariotti]

vswr cable loss calculation

What do yuo mean for "cable loss"? Perhaps the loss due to cable dissipation? The loss due to power dissipation is the which is quoted on the catalogs and data sheets.

The disspipation loss is 1/Available Gain (change sign if dB).

With VNA, collect spar , then with a simulator compute Available Gain.

If you have a SNA, there is no accurate formulation , but for practical cases you may apply:

Gav ≈ |Insertion Gain|/(1-|BRL|) [ use numbers! Not dB!]
BRL = Return Loss binned in frequency over at least 1 ripple cycle.

To check the result, Gav vs freq, should be a function like:

Gav = C0 * Freq^0.5 + C1 * Freq

and C0, C1 should be constant over frequency.

Many cables industries, like Huber+Suhner give the C0, C1 coefficients for most coax cables.

 

[biff44]

vswr cable chart

I think I see what you are saying: "If there is a large standing wave on the cable, you think that the power loss will be greater since there is a large voltage point somewhere out there and at that point V^2/R, where R is cable resistance, will be larger."

But I think you may be thinking incorrectly. You are sending power down the cable to the load. At the load, some power goes in, and some power reflects off (back to the source). So the power that goes into the load is what you are trying to trasfer. It does not matter at all if the reflected power then heats up the cable on its return (like in a big standing wave), or if it is absorbed in the source somehow. In either case, the same forward travelling wave power was transfered into the load. In other words, the forward travelling and reverse travelling waves are independent, so the fact that there is a big reverse travelling wave does not alter the forward travelling wave in any way. The same forward travelling wave power is incident on the load. The cable loss to the forward travelling wave is independent of if there is a big reverse travelling wave (Standing wave, or high VSWR) or not.

So, it seems to me that the cable loss measured in a matched measuring system is the SAME loss that you will experience in supplying power to the input of the load.

 

[jkkpohl]

I understand your perspective too because I once went down those lines and then doubted my logic. Isn't your proposal only valid for the very first wave to reach the load? Afterwards, and soon afterwards, the standing wave is established and these losses would be seen for all the furture incident waves. After all the standing waves are due to the constructive and destructive interference of the combining waves not just the reflected wave.

Sorry, but I am playing devils advocate a bit just so I can be sure of the analysis for this problem.

 

[biff44]

You can have tripple travel, etc. But it is a very unusual condition. Most people have a source that is matched to the transmission line, but have a load (like an antenna) that is not well matched. In that case there are only two waves travelling on the line, the forward toward the load, and the reverse reflecting off of the load. The reverse wave is absorbed when it finally reaches the matched source. The standing wave is made up of only the forward and reflected waves, and where they constructively and destructively add up.

Even if you had a unmatched source, it has to have a reflection coefficient of much less than 1 (or it would not have been a very good source), so the tripple travel wave would be small, if not zero.

 

[xxargs]

You can have tripple travel, etc. But it is a very unusual condition. Most people have a source that is matched to the transmission line, but have a load (like an antenna) that is not well matched. In that case there are only two waves travelling on the line, the forward toward the load, and the reverse reflecting off of the load. The reverse wave is absorbed when it finally reaches the matched source. The standing wave is made up of only the forward and reflected waves, and where they constructively and destructively add up.

Even if you had a unmatched source, it has to have a reflection coefficient of much less than 1 (or it would not have been a very good source), so the tripple travel wave would be small, if not zero.

I try to make simulation on vipec to see how sensitivity this is.

for example expensive coupler with 40 dB directivity is meaningless if you not have very precise cable impedanse connected to coupler - common cable with 48 - 52 Ohm precision (depend of batch) give only 33 dB directivity on ideal coupler in worst case. You must use cable and connectors with 50 ± 1 Ohm or better if you want 40 dB directivity in (ideal) coupler

RF-source impedance must be very close to 50 Ohm or you read differ value of reflected wave depend of phase angle of reflected power.

try this in vipec - is easie to see itself in simulator compare to try decribe this in bad english ;-)

---

Download free vipec linear network simulator (windows version)

https://sourceforge.net/project/showfiles.php?group_id=19792

Select ViPEC-3.1.3.zip, download and install (packing up in directory and starting program in directory)

load DIR_COUPLER.zip here and try and poke around with differ value on source/loads and cable impedance and length and see result.


coupler is a ideal 20 dB directive coupler
S31 is 'measured' reflected power from load (give -20 dB in case of total reflection)
port 1 is source generator - setting to 48 Ohm for non perfect generator
port 2 is load - here setting to 50000 Ohm for simulate open end
port 3 and 4 is terminator to make balance in coupler.

Using 48 or 52 Ohm cable in output of coupler, 'downgrade' coupler to max 33 dB directivity, if you want 40 dB directivity - try 49 or 51 Ohm cable to coupler output.
(simulate real life situation...)

cable 'lenin' is cable between source and coupler
cable 'len' is long cable between coupler to load

use tools/tune in meny to make real time update of chart with adjusting cable length.

resistor value 131.6 (R3) resp. 71.63 (R1) in PI-bridge is a 15 dB attenuator to simulate long lossy cable. (Vipec have no model of lossy transmissions line depend of length)

and new value of PI-pad can calculate from:

gamma = wanted attenuate in dB / 8.686 (dB converted here to Neper)

R3 = Z0 * sinh(gamma)
R1 = Z0 / tanh(gamma/2)
Z0 is wanted port impedance (here 50 Ohm, but try same impedance as used cable, 48 Ohm if you try 48 Ohm cable)

(hyperbolic equations is very nice compare to more clupsy arithmetic equation to calculate resistance value for PI-attenuator)

---

Vipec is very simple linear simulator and cannot replace very expensive EADS, ansoft etc. products , but for simple test and small network, this working pretty well and is free to use.

 

[brumi]

IF WHAT YOU ARE LOOKING IS TO MEASURE THE SWR , YOU NEED A DIRECTIONAL COUPLER.
You connect the cable to the signal source and at the end of the cable connect a terminator (resistive load with the impedance of the signal source)(ex 50 ohm).
Then measure the power reflected and have the SWR.

Added after 2 minutes:

Sorry I forgot said that to measure the cable loss you must connect at the end of the cable a short circuit terminator.
You measure the direct power and the reflected power and the diference between both is the cable loss.