clanon

Try this : Luneburg lens the Dielectric Constants should be changing from low outside to hugest in the core (twisting the Beam towards the center) https://www.analogictips.com/wp-content/uploads/2023/08/WTWH_Luneburg_lens_Pt3_Fig4.png also flat disc https://www.sciencedirect.com/science/article/pii/S2211379724000329

lo mas fácil es tener un cable (corto) de 50 ohms desde la antenna hacia la wlan y el conector hembra a rosca sin soldar dentro del PCB Easiest way is a coax from antenna (SHORTEST POSSIBLE) towards the usb wlan card and the SMA female attached to the card (and avoid soldering inside the card) incluso el conector roscado al reflector Disc seria lo mas corto y fuerte (el ultimo disco y mayor diámetro puede estar directo al exterior del coaxial)

Most people use it like this picture (the aux antenna could give problems) La mayoría de la gent lo usa de esta manera...la segunda antenna interfiere y no ayuda en nada...

i left it long ago...something is really OFF TWO different antennas (2.4 and 5.8 apparently) glued together and tryn to be fed at the same time...(i'll take a look again when I CAN) FEED NETWORK is MISSING...IMPROVISED and STUCK there THERE ARE easier and better collineal sector arrays..

Directors EXACT size might matter and distances between them...Should look like this...

Proximity feed (quarter wave Lambda) Widens the bandwidth impedance...a little I didn't like mine , if YOU want to tweak it , WORK over the feed tab size and distance to optimize the impedance bandwidth

A little more bandwidth with the Proximity FEED TAB (instead of direct contact)But yagi bandwith is real limited...an NEW whole band Bow tie LPDA would have better bandwidth...

Company published VSWR

I'm goin to replace the squares with bowtie triangles to check if it gets better... do you have the N type connector posted somewhere...?

Jomy ,could you post the aproxímate dimensions...I wanna try and make it wider bandwidth)

I'll see this weekend what can i do...(i don't like Yagis tho) Gain without LENGTH is not Achievable...or a 4 element ARRAY (short) feeding network should be hard to build...Did you EVER try a disk Yagi inside a TUBE...?

STL (i have cst 2019)

Don't let the AMOUNT of Signals Fools you ,You can see in the video what FR4 does to a small signal Wlan card little power output (it eats it up) Right HERE can't connect to AP...LOTS of AP doesn't mean Good Connection...Good RSSI NEITHER , PACKET LOSS RX Quality is more important BADLY DONE TEST

the ONLY advantage of Dielectrics...is SIZE reduction...and looks...ALL metal is the best option...