So I recently bought two 2,4GHz Yagis from different sellers on AliExpress for ~12€ each.
The first one was actually not for me but for a friend who is trying to extend the range of his WiFi-dependent drone. Mine serves primary as a range extender for my home WiFi so that I have access to the music server when BBQ'ing in the garden (and my smartphone-addicted friends can use Spotify or WhatsApp or whatever).
Both antennas had EXACTLY the same dodgy assembly 'mistakes' documented here (The first one was fixed in the same way, but I didn't document it from start to finish).
There seem to be a lot of similar antennas on eBay, AliExpress etc. and these could be common problems with many of them not working as intended, so I decided to document my findings.
Upon the first test via a Ralink USB WiFi-Dongle I noticed that the reception was actually worse than the integrated WiFi of my Laptop.
The length and spacing of the director rods / tubes also seems a little too even - a 'properly' calculated Yagi should perhaps have some more variation, for example as explained by this excellent tutorial here:
Investigating further with "horst", a Linux WLAN monitoring and troubleshooting tool, I discovered that the antenna had no directionality at all, there was poor reception and signal strength regardless of the way it was pointed in regard to my (and my neighbor's) router / access points. It could only pick up the closest access point and displayed a very modest signal strength.
First thing I noticed when checking for continuity on the antenna cable was that the antenna frame was connected to the shielding AND the core of the coax wire (!).
According to Wikipedia, the antenna frame and the director/reflector rods of a Yagi should NOT be electrically connected at all.
The active element CAN be a "folded dipole", that means the core and the shielding of the coax are connected to the opposite ends of a loop-style antenna element like the one used in this Yagi, so continuity between shielding and core alone would be intentional.
Next I disassembled the antenna / pulled off the active antenna element (you can actually "wiggle" it off to the front of the antenna - when fixing the first antenna I unnecessarily removed the reflector rod and mounting plate at the back to get access).
As I suspected (because of the location and length of the mounting screws) there was continuity between the metal sleeve / 'cable gland' of the active antenna element and the mounting screw towards the back end of the antenna.
Somewhat like the 'long screw damage' on Apple products... This screw was ~0.7mm too long.
Next I removed the hot glue used to 'seal' the active antenna element and found a piece of black wire embedded in the hot glue. On the first antenna the wire was actually soldered in as a bridge (!) between both ends of the folded dipole, on the second antenna one end of the wire seemed to be loose.
On the first antenna, the screw was longer, actually punched a hole in the sleeve and shorted the coax inside...