DK7ZB yagi. Insulate or not, that is the question

Magic band discussion - antennas, propagation, operating, etc
Forum rules
By registering any account on, you are accepting the Terms Of Use ("TOU").
The TOU may be viewed at:
Forum Diehard
Posts: 1675
Joined: Mon Jan 19, 2009 1:16 pm

Re: DK7ZB yagi. Insulate or not, that is the question

Post by VK2GOM » Tue Jan 05, 2010 8:17 pm

Hi Andy,

I too have seen this rounded cornered Aluminium listed, but the radiusing was on the inside of the tube, not the outside.

Edit: As a sort of PS to this, I just looked in the Capral catalog (big aluminium distributors) and they show tube with a rounded outside too. Interesting :)

73 - Rob VK2GOM / G0MOH

Forum Novice
Posts: 14
Joined: Mon Jan 04, 2010 9:11 pm

Re: DK7ZB yagi. Insulate or not, that is the question

Post by G0KSC » Wed Jan 06, 2010 12:40 am

A note or two on matching.

While this will be controversial, I have written an article on antenna matching which will be published in DUBUS 1 2010. Within the article I discuss the various methods of antenna matching the low impedance feedpoint to suit 50 Ohm coax connection. Also discussed is the fact that matching arrangements are added AFTER model. Even optimisation programs such as YO which model matching units only model the effect on the feed impedance, not the effect on the antenna pattern. In addition, as well as the impact the physical presence of a matching device has on the antennas pattern is the hole it leaves in the current at the most important part of the antenna which ultimatley will lead to a degraduation in performance. Below is a link to a an image of a 4 element Yagi fitted with a Hairpin (the most ‘lossy’ of the bunch). The pink lines denote the current within the element lengths, note the massive drop.

Having NEC4 I am able to accurately model matching devices and their effect. This has not been possible to do before (accurately) due to the inability of mininec and nec2 to handle sharp bends and stepped radius loops. Within the article a 7el 2M antenna is designed which is then fitted with a hairpin. 10dB F/B was lost in addition to .5dB forward gain. Much work had to be done in reshaping the conventional hairpin and changing element sizes and spcaing in order to get anywhere close to an acceptable level of performance.

Without re-hashing the article and all details, the conclusion drawn is that the Bazooka stub match (of which the DK7ZB match is a variant) is the least pattern and performance impacting matching arrangement that can be used for a Yagi. Provided a good quality, RF grade coax is used and the intended impedance of the antenna is maintained for the bandwidth the antenna is intended to be used across. The later point is important because if the Yagi has a 'ski slope' impedance (28Ohm antenna with an impedance of 35Ohm at one end and 18Ohm at the other end of its intended usable bandwidth) then power transfer across the transformer to the antenna will become inefficient once antenna deviates far from 28 Ohms.

Just a few thoughts to be getting on with :?:

Justin G0KSC

User avatar
Forum Diehard
Posts: 493
Joined: Fri Feb 08, 2008 10:29 pm
Location: Bay of Islands

Re: DK7ZB yagi. Insulate or not, that is the question

Post by ZL1RS » Sat Jan 09, 2010 4:41 am

Following discussions with Justin G0KSC it seems that these "EA4TX" insulators contain carbon and can influence the element "tuning", thus requiring a correction factor.

Carbon (lamp black) is added to plastics to get the nice looking black colour and improve the plastic's UV properties, but it also has effects at RF. This has been mentioned in other Forum threads ... i.e. problems with DK7ZB yagis being traced back to the plastic boxes used to mount the driven element and house the feedpoint.

This also explains a discrepancy in the "tuning" of a narrow band 9 element yagi I built recently. To cut a long story short, I built the antenna with free space element lengths to test how much the boom and insulators moved the SWR curve. Once I had the "as-built frequency offset" measurements, I applied the various known correction factors for boom proximity and element type and still "came up short" ... the only thing left was the element mounting material.

To fix the offset I simply scaled the element lengths according to the frequency error and cut another set. The final result was OK ... SWR curve and radiation pattern now following the computer modeling's curves.

These corrected elements were tried with different mounting insulators on the same boom and the SWR dip frequency moved significantly. Insulators do have an effect.

Somewhere on DK7ZB's web site there is a reference to tuning his antennas (probably in relation to the narrow band 12.5 ohm series?). He recommends making the elements a few mm too long, checking the SWR curve and then trimming all the element lengths in 1mm steps until the SWR curve meets that predicted by the antenna modeling software. The antenna's other properties should then be very close.

Another approach is to choose a very wideband yagi design so small discrepancies caused by construction (e.g. the insulator effect) will not unduly upset the final product.

Note that the insulator effect will be worse on higher bands where the length of the insulator in contact with the element is longer in terms of wavelength.

Posted in the interests of less hair pulling amongst home brew antenna builders :mrgreen:
Bob, ZL1RS in the Bay of Islands at RF64vs

Post Reply