Modular is good!

[ai8p]

Bob,

I am characterizing feedlines. The way this seems to work is that I short the end of the line and then tell it to measure the feedline. I think at that point it measures electrical length, loss, etc. Then it asks me to open the end of the line and hit OK for it to measure the line impedance. Sorry if I got that backwards, I don't have the manual in front of me

Now, if I want to characterize the line at different frequencies I must repeat this. For instance, I might want to get the loss at 1.8, 3.5, 7.0, 1.1, 14.0, 21.0, and 28.5 MHz.

Right now, in order to do this I have to do the following:

Set freq to 1.8
Climb tower to short end of line
Climb down tower and return to shack
Hit OK
Climb tower to open end of line
Climb down tower and return to shack
Hit OK

Set freq to 3.5
Climb tower to short end of line
Climb down tower and return to shack
Hit OK
Climb tower to open end of line
Climb down tower and return to shack
Hit OK

.................Repeat for each freq.

I have 2 feature requests:
1. After it reads the initial shorted (or open) termination, give me an option to get all the data it already knows without having to reverse the end connection and continue to get the other additional data. Maybe I only want the first set of data, and I don't want to change the termination to get additional data that I don't want.

2. Give me an option to store the data it obtained from the initial shorted termination and then reload it later to evaluate the alternate termination. In this case the procedure would be:

Set freq to 1.8
Climb tower to short end of line
Climb down tower and return to shack
Hit OK
Save shorted results for 1.8MHZ

Set freq to 3.5
Hit OK
Save shorted results for 3.5MHZ

............. Repeat for all freq of interest

Then:

Climb tower to open end of line
Climb down tower and return to shack
Set freq to 1.8
Load shorted data for 1.8 freq
Evaluate for open termination and finish analysis

Set freq to 3.5
Load shorted data for 3.5 freq
Evaluate for open termination and finish analysis

........ Repeat for all freq of interest

I'm not sure I explained this well enough for it to make sense, but I'll throw it out there and see what people think!

Thanks for listening

de Dennis, AI8P

[Bob]

Hi Dennis,

Although climbing the tower several times might be good exercise, you can probably get the information you need with two trips. One to disconnect the feedline from the antenna and another trip to restore it.

Disconnect the feedline at the antenna and run "Distance to fault". That will give you the impedance and the length of the line.
Note: if the line is balanced, like ladder line, use a 1:1 balun at the transmitter end to convert it to unbalanced.

You can also put a resistor across the far end of the line and measure the swr to get the line impedance. One trick to save a trip up/down the tower is to connect the resistor with a very small gauge wire. The AIM can still take it's measurements with no problem. After getting the data with the resistor load, connect the transmission line to your transmitter and key it briefly to burn the wire open. Then make measurements with the line open circuited. (This assumes your xmtr can stand an large swr for a few seconds).

With the line open, do a regular scan over the desired frequency range. You can save the scan data for comparison later in case you want to check for a change in the characteristics of the line. Then move the cursor along the scan data and read the cable loss at any frequency.

I'm sure others have noticed it's awkward to make an adjustment at the antenna and then have to go back inside just to press a key on the keyboard. I haven't looked at the remote control possibilities in detail but perhaps someone has or would like to. Here's one page with some ideas:
www.pcremotecontrol.com/

Other links popup if you Google on "pc remote control".

-- 73/ Bob

[ai8p]

Bob,

As I'm sure you must have surmised, I'm not an EE, so I'm feeling my way around this topic. So if you will bear with me, I will ask some really stupid questions.

For the Distance To Fault.

I assume I can have the end either shorted or open. Generally I prefer shorted when possible, because shorted is shorted, but open is whatever, depending on capacitance, etc of your end condition. Also, I do have a run of window line that I want to measure - I'm unclear why I should put a balun on the analyzer end instead of just connecting the ends of the wires to the binding post adapter. I actually have a balun between my coax and window line, and I can measure at the input of the balun, but I just am trying to figure out why. I also already have the capability to calibrate the meter at the binding posts, but I will need to solder up some coax couplers with short, and known resistor to calibrate for the input to the balun.

I presume from this that I will get electrical length and characteristic impedance of the line. I am pretty leary of putting a resistor across the far end and trying to burn the wire up with my transmitter, so I think I'll skip that!

One of the things I am most interested in is the matched line loss at various frequencies (I'm actually trying to program losses into my antenna modeling software to get the model to match the AIM readings). As I understand your note, I can just scan the line with the end open and the loss will be available in the graph for each frequency. For my coax this is straightforward. Do I assume correctly that for my window line I again need to measure at the input of the balun? Again showing my ignorance, but if I am looking for the matched line (SWR=1) line loss, do I need to terminate it in a resistor equal to the characteristic line impedance to measure loss, or is the open end measurement going to give me that.

Again, I'm sorry for the elementary questions - perhaps you need a quiz that people need to pass before they would be qualified to purchase your meter. Hi Hi.

Thanks for you help

de Dennis, AI8P

Apr 25, 2008 14:41:50 GMT -5 Bob said:

Hi Dennis,

Although climbing the tower several times might be good exercise, you can probably get the information you need with two trips. One to disconnect the feedline from the antenna and another trip to restore it.

Disconnect the feedline at the antenna and run "Distance to fault". That will give you the impedance and the length of the line.
Note: if the line is balanced, like ladder line, use a 1:1 balun at the transmitter end to convert it to unbalanced.

You can also put a resistor across the far end of the line and measure the swr to get the line impedance. One trick to save a trip up/down the tower is to connect the resistor with a very small gauge wire. The AIM can still take it's measurements with no problem. After getting the data with the resistor load, connect the transmission line to your transmitter and key it briefly to burn the wire open. Then make measurements with the line open circuited. (This assumes your xmtr can stand an large swr for a few seconds).

With the line open, do a regular scan over the desired frequency range. You can save the scan data for comparison later in case you want to check for a change in the characteristics of the line. Then move the cursor along the scan data and read the cable loss at any frequency.

I'm sure others have noticed it's awkward to make an adjustment at the antenna and then have to go back inside just to press a key on the keyboard. I haven't looked at the remote control possibilities in detail but perhaps someone has or would like to. Here's one page with some ideas:
www.pcremotecontrol.com/

Other links popup if you Google on "pc remote control".

-- 73/ Bob

[w0qe]

Dennis,

Don't waste your time making too many measurements.

Over the HF and low VHF frequency range the matched loss of coax (feedline) is primarily determined by the characteristic impedance of the coax (feedline) and the size of the center conductor (both conductors if balanced line). The matched loss is very very closely proportional to the square root of the ratio of the frequencies in dB. For example measure the loss at 10 MHz and the loss at 40MHz will be twice as much in dB per unit length and half as much at 2.5MHz.

Times microwave has a very good coax loss calculator at
www.timesmicrowave.com/cgi-bin/calculate.pl
which shows the loss being proportional to the square root of frequency for a large number of types off coax.

To measure the characteristic impedance of coax (any feedline) take a reasonable length (>1/8 wavelength) and measure the impedance with the far end open and then again with it shorted. Take the two impedances (complex numbers) and multiply them together and then take the square root which will give the characteristic impedance (independent of line loss). This test requires careful calibration of the AIM4170 and is better done when the line is not an exact multiple of 1/4 wavelength which produces very high and very close to zero impedances. Don't be surprised if the answer is something line 50.5 + j.5. Very few coax cables have a characteristic impedance that is perfectly resistive. For example RG8-X is about 50.2 - j0.7 and RG-213 is 50.0 - j0.39.

73,
Larry, W0QE

[w0qe]

Dennis,

One more thing. With every copy of the ARRL Antenna Book comes a program by Dean Straw, N6BV called TLW. This program does many of the calculations you are interested in and is an excellent way to understand transmission lines.

73,
Larry, W0QE

[ai8p]

Larry,

Thanks very much for those tips. I have played with TLW quite a bit and actually have programmed up the Transmission Line formula from the Antenna Handbook to try to match it. I am very close to agreeing with the TLW results for my specific setup. I'm actually about 0.5dB low in my loss calcs compared to TLW for a setup that has about 9dB total loss in the line. However, for higher frequencies, I am not predicting enough loss. I am running a model with just the antenna, then importing the feedpoint impedance into Excel and doing the TL loss calcs for both the window line and the coax within Excel. I am using the loss factors given in TLW for typical lines, and I want to measure the loss for my specific lines to see if that helps close the gap. I also thought since I have the meter I would measure the loss at various frequencies and see if the thumbrule for loss at different frequencies works well for this setup. Once I have the model calibrated to produce a reasonable agreement with my measured values, then I can use the model with confidence to experiment with alternate antenna/feed line configurations to optimize my setup.

I would appreciate it if anyone would like to answer my specific questions about the measurement techniques.

Thanks for your support.

de Dennis, AI8P

[w0qe]

Dennis,

Yikes, 9dB of loss in the line is a lot. That is the same as 7/8 of the power being lost in the transmission line. I believe that your inconsistency in the loss may be due to mismatch loss. Try replacing every PL-259 connection in the system with a piece of 30 ohm transmission line that is 0.9" long and a VF of 1 and see if that changes things. If you use any kind of PL-259 barrel then you will need to use the distance of the barrel plus about .6" on each end. With a reasonable understanding of how transmission lines work and some care in making the measurements you should be able to get pretty close.

Also be careful with the ARRL Handbook formula for the additional loss in coax when the SWR is other than 1.0:1. The formula returns the correct average loss while TLW calculates the loss correctly. For example both 10+j0 and 250+j0 both represent SWRs of 5.0:1 as loads but given a 10 ft. piece of RG-58A coax at 7Mhz the coax connected to the 10+j0 load will incur an extra 0.388dB of loss but the same piece of coax connected to the 250+j0 load will only incur an additional loss of 0.017dB. However changing the length of the coax to 92.7 ft. both loads will produce additional losses of 1.247dB. This is because the current in the coax changes along the length but at HF the loss is highest per unit length where the current is highest.

This topic while interesting might be better served if you want to contact me directly since it really is not an AIM4170 issue. My QRZ email address would be fine.

73,
Larry, W0QE

[Bob]

Apr 28, 2008 7:16:59 GMT -5 ai8p said:

Bob,

Also, I do have a run of window line that I want to measure - I'm unclear why I should put a balun on the analyzer end instead of just connecting the ends of the wires to the binding post adapter. I actually have a balun between my coax and window line, and I can measure at the input of the balun, but I just am trying to figure out why. I also already have the capability to calibrate the meter at the binding posts, but I will need to solder up some coax couplers with short, and known resistor to calibrate for the input to the balun.

As I understand your note, I can just scan the line with the end open and the loss will be available in the graph for each frequency. For my coax this is straightforward. Do I assume correctly that for my window line I again need to measure at the input of the balun? Again showing my ignorance, but if I am looking for the matched line (SWR=1) line loss, do I need to terminate it in a resistor equal to the characteristic line impedance to measure loss, or is the open end measurement going to give me that.

Thanks for you help

de Dennis, AI8P

Apr 25, 2008 14:41:50 GMT -5 Bob said:

Hi Dennis,

By calibrating after the balun, you take into account imperfections in the balun itself. Baluns have some length of cable in their construction, so they can't be considered as ideal transformers. Their properties are frequency dependent.

To measure the cable loss, the far end has to be open or shorted so that the reflection coefficient at the end is unity. If you have a resistor there, then the resistor will absorb some of the energy and it will look like the cable loss is more than it really is. The AIM program assumes that whatever did not come back, must have been lost in the cable.

For working with complex numbers, I've found a free program called "SpeqMath" to be very useful:
www.speqmath.com

If you want to measure the impedance with a short and with a open and then take the square root of the product, this program is really nice.

73/ Bob

[dl4raj]

Hi Larry,

thanks for the link to the loss calculator,very useful.

Apr 28, 2008, 11:14am, w0qe wrote:...
>Very few coax cables have a characteristic impedance that is
>perfectly resistive. For example RG8-X is about 50.2 - j0.7 and
>RG-213 is 50.0 - j0.39.
....

There's actually no coax with perfect resistive impedance
because it would to have to be lossless.
As strange as it may seem but the resistive losses in transmission lines are the reason why their impedance not resistive but complex.

73
Clemens
DL4RAJ