Unexpected antenna impedance difference

[kellogs]

Hello,

Homemade BLUE bowtie antenna vs factory made BLACK whole board with the same bowtie antenna.

Over 300 - 330 MHz range I am seeing some ~20 change in reactance with the home made antenna and ~50 ohm with the factory made. Also the resistance is 7-8 times greater for the factory made antenna, also varying much more.

Sames:
- placement of the antenna
- calibration place, cables, balun and others related (*)

Diffs:
- calibration range: 300-330 for home made vs. 235-395 for factory made
- (*) connection to home made antenna through SMA connectors vs. direct soldering for the factory made antenna;
- calibrated with standards for home made vs. with wire / nothing / two 100 ohm resistors measured at 50.5 ohms with DMM

Is it normal to see such large measurement differences ?

Thank you

 

[FvM]

Can't recognize that both antennas are equivalent. Black antenna is connected through two cables and balun, blue antenna through single cable with asymmetrical connection. Would prefer a drawing instead of photo with blurry details.

 

[kellogs]

- (*) connection to home made antenna through SMA connectors vs. direct soldering for the factory made antenna;

same balun and two cables were used in both cases

 

[volker@muehlhaus]

same balun and two cables were used in both cases

Show a photo how you connected the two coax cables to your homemade antenna.

 

[vfone]

For me the two antennas looks pretty much the same, assuming the same dimensions and the same substrate Er.
In this situation, just feed the booth antennas in the same way, and you may get the same results.

 

[FvM]

A basic bow-tie antenna with asymmetrical feedline connection causes the feedline to radiate, therefore construction details matter.

 

[kellogs]

@volker@muehlhaus I have not taken any photos of the home made blue antenna when the balun was still connected to it, but it is not hard to picture: the same two cables seen in the black antenna photo had had their center conductors soldered to a SMA connector that threaded into the blue bowtie's SMA connector.

@vfone @FvM The blue, home made antenna is a bit asymmetric between its sma connector and where its copper pour begins on the PCB, while the factory made black antenna is symmetrical. And still, the better measurements are with the home made antenna...

 

[FvM]

O.k. feedline geometry is partly different. Without sheet wave filter (ferrite tube) near antenna, coax cable radiates and affects antenna impedance considerably. It's just expectable that both geometries involve different impedance.

 

[kellogs]

Do the two coax cables used like this not form a differential pair, which is not supposed to radiate much ?

Hmm, one more difference I have neglected: factory made antenna (the worse one) has the coax sheets connected to GND plane of the PCB at a distance from the antenna. Oh, and also, there is no GND plane with the home made antenna. Let me cut that connection; done:

- resistance is still much greater than for home made, but it does not vary that much anymore
- reactance is still about the same, and still varies the same, maybe a tad less

Maybe the ground pour is affecting the impedance so greatly ? I have just used this calculator, adjusted wavelength for PCB epsilon_r and thought that 2 cm was well withing the far field... the GND pours perimeters are drawn in grey in the picture attached.

 

[FvM]

Do the two coax cables used like this not form a differential pair, which is not supposed to radiate much ?

Yes, if connected symmetrically. Unfortunately the connection details can't be decoded due to blurry photo. "Blue" antenna has apparently asymmetrical coax connection, similar to the drawing in post #6, so the cable does radiate.

It's also not obvious that used baluns are optimal. I wonder what kind of balun is soldered to the PCB in your latest post.

 

[kellogs]

Attaching a clearer photo.

Blue antenna does not have asymmetrical coax connection and is unlike the drawing in post #6; it uses the same pair of two coaxes, not just one coax, and the same balun. It does, however, have asymmetry in the PCB traces from its SMA connector on the PCB towards the antenna elements. Also, the measurements are close to the expected impedance with this antenna, indicating that the cables do not radiate.

The balun (150 kHz - 500 MHz): https://www.aliexpress.com/item/1005006432112776.html

LE: I have cut away the PCB part with component footprints on one side and GND plane on the other; getting the same readings.

 

[FvM]

May be I'm misreading the photo. I'm seeing SMA ground pin connected to left bow-tie element. Also no second SMA visible.

 

[kellogs]

Those are correct observations. Point being ? Maybe this addresses it:

@volker@muehlhaus I have not taken any photos of the home made blue antenna when the balun was still connected to it, but it is not hard to picture: the same two cables seen in the black antenna photo had had their center conductors soldered to a SMA connector that threaded into the blue bowtie's SMA connector.

 

[volker@muehlhaus]

but it is not hard to picture: the same two cables seen in the black antenna photo had had their center conductors soldered to a SMA connector that threaded into the blue bowtie's SMA connector.

From what I understand, you are breaking symmetry.
And from what I can see, your brute force soldering is not really compatible with the fine pitch detail at the antenna feeds (commercial antenna).

From my experience, you need to pay more attention to the detail at the frequencies. You should work like a watch maker here, not like a plumber