Month: March 2014

Now with my antenna cable, 20m of 5D-2V.

First, both channels at the dummy load.

Then, the delay is measured to be -42.80nS at 7.026MHz (T=142.32nS), which means 99.52nS of delay.

Since the physical length of the cable is 20m, the velocity factor is computed to be 0.67, which is quite a reasonable value.

The impedance bridge shows that the SWR value is almost 1.0, because the 20m cable is terminated with a dummy load.

The voltage difference at the impedance bridge and at the dummy load is due to the cable loss of 20m.

The measured loss is around 0.61dB. Note that the loss of 5D-2V cables at 10MHz is 26dB/km or 0.52dB/20m.

If you remove the dummy load and terminate the cable with an open circuit, the impedance bridge shows:

Note: Something is wrong with the green trace. It should be ch1-ch2 signal, but obviously it is not.

gnuplot> load "gnuplot.txt"
Freq [MHz]=7.026
V1=2.426
V2=1.344
Cursor 1=-2.72e-08
Cursor 2=0.0

vratio=0.553998351195383
phase1 [deg]=-68.798592
phase2 [deg]=0.0

abs(gamma)=0.951950606187447
swr=40.623834169527
cz={1.33775338471275, -14.7339917147814}

With the cable length of 20m and the velocity factor of 0.67, we have:

This shows an open circuit at the far end of the cable.

Cable length measurement to check the velocity factor. The RF signal from the transmitter comes from the left, and the electrical length of the cable (cyan) is determined by measuring the signal delay between ch1 (red) and ch2 (yellow).

This is when both ch1 and ch2 probes are at the dummy load. The frequency of the RF signal is 7.026MHz, and there is no delay between the two channels.

This is when the ch1 probe is moved to the left end of the cable under test (cyan) as is shown in the first figure.

The delay of 11.40nS is observed.

The RF signal is now QSYed to 14.052MHz. Of course, the delay is the same whatever the frequency is.

This is with another cable with the same specifications. The delay is now 11.60nS.

Since the delay of 1nS corresponds to the 0.3m of the electrical length, the length of the cable is 3.42m. Since the physical length of the cable is 2.0m, the velocity factor becomes 0.58, which is somewhat different from the the velocity factor of 0.67 of RG-58U cables.

I noticed that something is wrong with my 40m band dipole with stub match. The SWR value is now almost 3.0.

gnuplot> load "gnuplot.txt"
Freq [MHz]=7.026
V1=7.08
V2=3.68
Cursor 1=5e-09
Cursor 2=0.0

vratio=0.519774011299435
phase1 [deg]=12.6468
phase2 [deg]=0.0

abs(gamma)=0.505804557491017
swr=3.04698187795135
cz={16.5995525105867, 5.07690836540991}

The measured impedance of cz={44.6133821915143, 24.2820051981698}, the red point, corresponds to Zant(with stub)=35.252-j16.472 ohm, the green point, if you consider the cable (5D-2V) length of 20m.

Without the stub match, Zant=27.182-j40.050. Therefore, it is presumed that the stub works something like this way.

Now yet another single stub match for my new 40m band antenna. Using 5C-FV (z0=75 ohm, and the velocity factor is assumed to be 0.75), the parameters are d1=940.9mm and d2=2839.3mm, and we supposed to attain Z=50.379-j0.261 ohm.

This is with a stub match. Looks nice, isn’t it?

gnuplot> load "gnuplot.txt"
Freq [MHz]=7.026
V1=8.0
V2=8.32
Cursor 1=5.6e-09
Cursor 2=0.0

vratio=1.04
phase1 [deg]=14.164416
phase2 [deg]=0.0

abs(gamma)=0.254631445240079
swr=1.68323634962598
cz={44.6133821915143, 24.2820051981698}

A dummy load is connected to check how the stub is working. The black coax cable forms the stub section, and the gray cable goes to the transmitter.

gnuplot> load "gnuplot.txt"
Freq [MHz]=7.026
V1=2.208
V2=2.59
Cursor 1=7e-09
Cursor 2=0.0

vratio=1.17300724637681
phase1 [deg]=17.70552
phase2 [deg]=0.0

abs(gamma)=0.375575673213821
swr=2.20295016418356
cz={47.3942331949417, 39.368039298651}

The current length of the stub section using 5C-FV is around 3750mm, and there is a cable of length 1000mm between the impedance bridge and the stub. Assuming the velocity factor of 0.75 for 5C-FV, the expected impedance measured by the impedance bridge is 41.731+j32.604 ohm, which is in good agreement with the cz=47.394+j39.368 ohm.

A stub (the section in blue) is usually placed near the antenna (orange) to minimize the cable loss. In other words, the parameter d1 is chosen to be the minimum from the set of possible valuse of d1. But sometimes it happens that you have your stub near the transmitter (cyan), that is in your shack.

This is a stub section in such cases. The transmitter is on your left side, and the antenna on your right side. (Actually, it can be reversed, because there is no direction.) As you can see, the end of the stub section is short-circuited, but it can be open if you adjust the length d1 properly.

The mean lifetime of my simple wire antennas is relatively short, and my folded weirdly antenna for the 40m band was no exception.
http://spinorlab.wordpress.com/2013/09/19/folded-weirdly/

Last weekend I re-installed the element with supposedly almost the same configuration, but this time with each leg of the element to be 10m long.

With the usual measurement it is observed that at 7026kHz CH1=2.943V, CH2=2.638V, and CH2-delay=+12.60nS.

gnuplot> load "gnuplot.txt"
Freq [MHz]=7.026
V1=2.943
V2=2.638
Cursor 1=1.26e-08
Cursor 2=0.0

vratio=0.896364254162419
phase1 [deg]=31.869936
phase2 [deg]=0.0

abs(gamma)=0.530090794923996
swr=3.25614135325678
cz={20.4433917939881, 26.9131366419314}

With a Smith chart and a physical cable length of 20m (remember the velocity factor of 0.66), you start from a red point (Zdut=20.443+j26.913), rotate towards load (not towards generator!), and end with a green point (Zant=27.182-j40.050), which shows the radiation impedance of the antenna.

Since the Arduino Leonardo has only 32kB (with 4 KB used for the bootloader) of flash memory for storing its program, not all the features of K3NG keyer can be enabled at the same time. For example, features serial Command Line Interface and Command Mode can not be compiled and downloaded simultaneously.

It is possible to prepare different versions of programs and download one of them depending on your necessity. If you want the serial Command Line Interface feature to be enabled, that means you have a serial communication with your Arduino board, in which case downloading a program will give you only a minor overhead.

There is a type named Arduino Mega 2560, which has a flash memory of 256kB, large enough to include all the features, but naturally the board is more expensive than the Leonardo.