[amsat-bb] Homebrew LFA yagi problems

[Hans BX2ABT]

I'm new to yagi building and so far have made a four element test LFA 
yagi for 70 cm using a wooden boom and thick copper wire. Worked as 
expected. I scaled it up and build a 10 element version very closely to 
the specs of the InnovAntenna version (thanks KB1PVH for the help 
there). You can find an photo here: 
http://bx2abt.com/main/data/_uploaded/image/20180924-yagi.jpg The main 
differences are the tubing (InnovAntenna uses 1/2", I use 8 mm) and the 
width of the loop (InnovAntenna uses 46 mm, the smallest my tube bender 
can do is 68 mm). Apart from these two the InnovAntenna has through-boom 
elements and I use on-boom elements.

According to the lenght of my loop the resonant frequency is 436 MHz 
(300/436 = 68.8 cm). Using a HT and my Daiwa SWR meter I only get 
resonance (SWR 1:1) at 412 MHz. Making the loop shorter in length 
doesn't make much difference. I understand that the reflector and 
directors bring the high impedance of the loop down to the desired 50 
ohms, hence they have a noticable influence on the resonant frequency.

How to go on from here? What am I doing wrong?

My first thought was to start trimming the elements up to 95% of their 
current lengths (412/436*100 = 94.5%), then adjust the element spacing 
to 95% of the original distance.

Second thought: there is (still) a feedline mismatch and the coax throws 
off the resonant frequency. On the G0KSC website there is a design for a 
1:1 symetrical coax balun and talk about the Pawsey stub. Worth trying?

I would appreciate input from people who have successfully homebrew LFA 
yagis and who could point me in the right direction. Also if you have 
experience with modelling LFAs using MMANA I would like to hear from 
you. Cheers,

Hans
BX2ABT