G4CQM is the Amateur Radio callsign held by Derek Hilleard, previously licenced G8CIW from early 1969.
Self-employed for more than twenty years as a professional Antenna Design Engineer for the radio and telecommunications industry. Now retired Derek presents websites focusing on specialist VHF/UHF/SHF yagi designs for fellow Ham Radio enthusiasts... | Blog Archive | antennadesigner projects | waveantennas projects | Atlantic DX |
Visitors Today: Guests Online:
Two working Q-factor examples above, CQM9UX0 and XX15 yagis both designs by G4CQM
A forensic study of Q-factor and yagis by Dragoslav YU1AW is to be found here, definately worth reading!
However, a simple tool/speadsheet is needed to carry out analysis of your design. Most of my home computing is done with the Linux operating system, a distro version called Lubuntu 14.04 LTS. The default spreadsheet package that comes with this os is called Gnumeric. Speadsheets are commonly saved with a .xls file extension and Gnumeric is able to do this.
Using Gnumeric I have devised a simple light weight speadsheet which when used in conjunction with the free and powerful 4nec2 program will calculate yagi Q-factor...
After running a frequency sweep in 4nec2 a VSWR plot will be produced seen on the left above. Use the drop down menu to select first R-in (real) then X-in (imag). Save the two text files as R.txt and X.txt...
Open Gnumeric, copy from the text files (freq/R then freq/X in turn) and paste into the selected designated areas. This can be done without having to manually copy and paste each single value, may need some practice if you are not familiar with this process...
Gnumeric will now update and display Q and Average Q for your chosen design. By right clicking on the graph (Properties) it is possible to save (Save as image) this as a .png image directly to your desktop or any folder of choice, your done!
Download and test my two samples... G4CQMQFCQM9UX0.xls and G4CQMQFXX15.xls (MS Excel™ 97/2000/XP)
Warning: NEC-2/4 based software is not good for analysis of loops and folded dipole driven elements, it will display incorrect VSWR plots. As a result Q-factor will be inaccurate! I am very fortunate to have K6STI's AOP (Antenna Optimizer Professional) version software which has bent wire correction built in and can be turned on or off as required by the user. When carrying out examination of complex driven element yagis I have carried out a swept analysis in AOP and manually entered the R and X results into my Gnumeric spreadsheets to be certain of accurate results!
A guide based on field trials (last five years) and recent analysis (2014)...
The simple rule, the lower the 'Q' the better is stability. Yagi designs sporting a very steep exponential curve when plotted are not ideal, this dramatically raises the average 'Q'!
At 144MHz some of the designs reviewed clearly have potential stability issues!
↑ Best ↑
↓ Worst ↓
At 50MHz stability is not a problem for the majority of yagis reviewed so far, all within the very good to reasonable stability range! However, Reflectorless 50R2 may become unstable in certain conditions!
↑ Best ↑
↓ Worst ↓
Read the full story on Q-factor here!
The Optimized Wideband Antenna (OWA) Yagi was a result of work primarily by Jim Breakall WA3FET back in the mid nineties. Recently Justin G0KSC has highlighted a flaw in many yagi designs where desired attributes are not consistent with one another over the desired bandwidth.
During my own quest to perfect the 50-Ohm yagi I noticed that by producing several iterations of one particular design theme that it was possible to eventually identify an optimal design, enter the AOWA (Advanced OWA).
Earlier this year I was contacted by a fellow enthusiast in Finland who was looking for a 15 element 144MHz design using 16mm OD tube for the DE and 12mm OD tube for the parasitic elements.
XX15 AOWA is the result of my design process. Take a look at the Gain, F/R and F/B graphs pictured above. Even though the design was optimised for performance in the 144 - 145MHz range the desired parameters are almost flat lines between 143 - 146MHz. Meanwhile Q-factor for this design is low (Average Q = 31.31)! This is what makes an AOWA very special!
G4CQM is looking for a Belcom Liner 2 DX. Must be in good working condition. Can you help? If so please contact Derek Hilleard G4CQM, thanks!
Majsai Mihaly HA8CL decided earlier this year to build my CQM12C4E (special EU version) design which has a very low temperature (single yagi T_ant = 218.3 K). This design also benefits from low Q-factor (Average Q = 29.52)! Excellent job Majsai, thanks for sharing your great pictures!
Email received from Barry G4KCT... Clamped the 60ele on the tower this morning with the 2 X 58ELE Q loops and did a test with G8PNN 150km north of us, S9 signals of both antennas, the 60ele having slightly less base line noise and a couple of Deg more beam width. I still had 20ft of Ecoflex cable coiled up on the 60ele. Its now on the tower with the 13cms antenna, the horizontal spacers between the antennas are fibre glass tube. The weight loss cranking the tower back over made a difference. Been sat listening to GB3CSB AND GB3GHZ over lunch and no problem hearing them. Ordered some more mounts to start the Mk2 version. regards barry 4kct
No matter what length of coaxial feeder you are using this simple test will reveal if you have a problem with your yagi beam. It is important to remember that VSWR should be regarded as a guide only.
Textbooks refer to an infinate length of coaxial cable that is unterminated presenting its own characteristic impedance (Zo), this is due to the process of self termination. Meanwhile a finite length terminated in its own characteristic impedance (Zo) will deliver that impedance. In both cases there will be no reflected power when tested.
If your yagi beam is truly 50-Ohms without any reactance present you will see a 1:1 match i.e. no reflected power.
The test: By inserting a λ/2 then λ/4 section ahead of your VSWR meter you should see no or very little change in reflected power. If you observe a large change it is likely that your yagi impedance is not 50-Ohms and/or reactance is present. In particular the λ/4 section will transform and magnify any anomoly present. These sections must be an electrical λ/2 and λ/4 incorporating the cables velocity factor (Vf) at the frequency on test. Make up these two sections and keep them in your tool box!
Warning: Beware there is fake and sub standard coaxial cable on sale worldwide at present, be sure of what you buy!
Barry Firth G4KCT recently decided to build my 60 element 1296MHz FLOWA design. A major problem for designers of 23CM yagis is software accuracy at this high frequency. This is mainly due to the limited number of segments per λ/2 available because of wire diameters. Segments have to be larger than wire diameter. As a consequence some manual tweaking of the driver cell is required.
Construction tolerances are also tight, cutting of elements and their positions needs to be within 0.1mm. Barry used a drill block pictured above to locate his elements on the boom. As for the driven element Barry adopted the sliding mechanism shown above to fine tune. The driven element has modified solder tags soldered directly to and over the 3/16 Inch brass tubing. Note his short lead lengths connecting the coaxial dipole directly to the ECOFLEX 10 feeder.
Barry's dimensions are as follows: Ref at 0mm, DE at 31mm and D1 at 56mm. Reflector length 120mm and DE 117mm end to end with 5mm gap. Barry reports that his Bird Thruline now shows almost no reflected power!
Driven element now in enclosure a small electrical box...
Barry will be making a MK2 version with semi rigid RG402 cable fed from underneath, housing the DE in a suitable enclosure. Thanks to Barry for providing this level of detail and pictures!
Richard Mason G6HKS has just updated the PowAbeam Antennas website. There are several yagi bargains to be had, pictures of Richards 432MHz LDMOS amp project and YouTube clip of NC1I being heard on 70CMS.
Those of you who regularly visit this website will see a big change. My Atlantic Ocean DX page has been closed down for a while, time needed to think!
G4CQM Blog © 2014 Derek Hilleard G4CQM | Contact: g4cqm(at)live.co.uk | HTML5 CSS3