ba4tb
New Member
Posts: 5
|
Post by ba4tb on Apr 10, 2014 3:03:14 GMT -5
Hi Bob,
Can you explain how the procedure of measuring cable loss using AIM4170?
Thanks
Dale
|
|
g3txq
New Member
Posts: 34
|
Post by g3txq on Apr 11, 2014 9:33:54 GMT -5
Here's the most accurate technique of many I've tried: 1) Near end of the line connected to the AIM4170; short the far end of the line 2) Perform a scan with the AIM4170 3) Identify the frequencies where X=0 and R is a minimum; note the values of R For small losses: Cable Loss (dB) = 8.69 * R / Zo Or more generally: Cable Loss (dB) = 10 * Log[(Zo-R)/(Zo+R)] If you need a few more frequency points, repeat with the far end of the line open circuit. The explanatory maths is here: www.karinya.net/g3txq/wet_ll/tl_formulas.pdfYou can see the sort of accuracy achieved using some 300 Ohm window line in the very last chart here: www.karinya.net/g3txq/wet_ll/Usually, an assumed value for Zo is adequate. If you need more accuracy use Zo=SQRT(Zoc.Zsc) from the measurements, but choose a frequency where the line is an odd multiple of 1/8 wavelength long; that will be frequencies where Xoc=-Xsc. Hope that helps, Steve G3TXQ
|
|
|
Post by Bob on Apr 11, 2014 12:04:23 GMT -5
Hi Dale,
Do a standard cal of the AIM with the cal loads at the front panel RF connector. Enable the Return Loss parameter graph ( Setup -> Plot Parameters -> Return Loss). Do a scan over the desired frequency range with the cable open or shorted at the far end. As you move the vertical (blue) cursor, the cable loss at each frequency will be displayed in the data on the right side of the graph.
This is numerically equal to one-half the return loss because the wave travels down and back through the cable, thus experiencing the cable loss two times. The AIM measures return loss directly and then divides that by two to get the cable loss.
73/Bob
|
|
g3txq
New Member
Posts: 34
|
Post by g3txq on Apr 12, 2014 10:13:18 GMT -5
If you use the method Bob suggested, you might want to perform one scan with the line short-circuit, and then re-scan with the line open so that the results are over-layed on screen. You'll likely see some very significant differences in cable loss reported between the two cases! You'll find that the results for the two cases are closest where the line impedance is purely resistive, and they differ most where the line impedance X/R ratio is highest. If you want best accuracy, restrict yourself to the purely resistive frequencies, where the line is a multiple of a quarter-wave long. What I described earlier is my "manual" version of that method; Bob's method saves you the maths - but both methods give the same answers. Edit: example here - www.karinya.net/g3txq/temp/aim_cable_loss.png50ft of RG58C cable. At 4.72MHz the cable loss is reported as 0.41dB with a S/C termination and 0.64dB with an O/C termination. Reported cable losses are identical where the line is a quarter-wave multiple long and the line impedance is purely resistive. 73, Steve G3TXQ
|
|
ba4tb
New Member
Posts: 5
|
Post by ba4tb on Apr 13, 2014 21:38:04 GMT -5
Thank you guys.
We have 3 cables(all 50 ohm) joint by 2 connectors,first one is 50m long,the second and the third are both 100m long. Are the results we measured of the length and cable loss only for the first 50m cable? How can we measure the entire cables?
73, Dale
|
|
g3txq
New Member
Posts: 34
|
Post by g3txq on Apr 14, 2014 2:40:35 GMT -5
You will be measuring the loss of the entire length.
Steve G3TXQ
|
|
|
Post by resonant on Apr 21, 2014 10:07:40 GMT -5
In doing Network Analyzer tests on transmission lines, I have found the most accurate way to get the loss of a line is to measure a specific frequency of interest, first with an open circuit, then a short circuit, then subtract the smaller figure from the larger and this gives you the loss in db for the cable's full length.
|
|